@article{Klein2020Modelling,
 author = {Klein, Nadja and Herwartz, Helmut and Kneib, Thomas},
 year = {2020},
 title = {Modelling regional patterns of inefficiency: A {Bayesian} approach to geoadditive panel stochastic frontier analysis with an application to cereal production in {England and Wales}},
 pages = {513--539},
 volume = {214},
 number = {2},
 issn = {03044076},
 journal = {Journal of Econometrics},
 doi = {10.1016/j.jeconom.2019.07.003},
 howpublished = {refereed}
}

@article{Pook2020Improving,
 abstract = {Imputation is one of the key steps in the preprocessing and quality control protocol of any genetic study. Most imputation algorithms were originally developed for the use in human genetics and thus are optimized for a high level of genetic diversity. Different versions of BEAGLE were evaluated on genetic datasets of doubled haploids of two European maize landraces, a commercial breeding line and a diversity panel in chicken, respectively, with different levels of genetic diversity and structure which can be taken into account in BEAGLE by parameter tuning. Especially for phasing BEAGLE 5.0 outperformed the newest version (5.1) which in turn also lead to improved imputation. Earlier versions were far more dependent on the adaption of parameters in all our tests. For all versions, the parameter ne (effective population size) had a major effect on the error rate for imputation of ungenotyped markers, reducing error rates by up to 98.5{\%}. Further improvement was obtained by tuning of the parameters affecting the structure of the haplotype cluster that is used to initialize the underlying Hidden Markov Model of BEAGLE. The number of markers with extremely high error rates for the maize datasets were more than halved by the use of a flint reference genome (F7, PE0075 etc.) instead of the commonly used B73. On average, error rates for imputation of ungenotyped markers were reduced by 8.5{\%} by excluding genetically distant individuals from the reference panel for the chicken diversity panel. To optimize imputation accuracy one has to find a balance between representing as much of the genetic diversity as possible while avoiding the introduction of noise by including genetically distant individuals.},
 author = {Pook, Torsten and Mayer, Manfred and Geibel, Johannes and Weigend, Steffen and Cavero, David and Schoen, Chris C. and Simianer, Henner},
 year = {2020},
 title = {Improving Imputation Quality in {BEAGLE} for Crop and Livestock Data},
 pages = {177--188},
 volume = {10},
 number = {1},
 journal = {G3: Genes|Genomes|Genetics},
 doi = {10.1534/g3.119.400798},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/31676508},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6945036},
 howpublished = {refereed}
}

Imputation is one of the key steps in the preprocessing and quality control protocol of any genetic study. Most imputation algorithms were originally developed for the use in human genetics and thus are optimized for a high level of genetic diversity. Different versions of BEAGLE were evaluated on genetic datasets of doubled haploids of two European maize landraces, a commercial breeding line and a diversity panel in chicken, respectively, with different levels of genetic diversity and structure which can be taken into account in BEAGLE by parameter tuning. Especially for phasing BEAGLE 5.0 outperformed the newest version (5.1) which in turn also lead to improved imputation. Earlier versions were far more dependent on the adaption of parameters in all our tests. For all versions, the parameter ne (effective population size) had a major effect on the error rate for imputation of ungenotyped markers, reducing error rates by up to 98.5%. Further improvement was obtained by tuning of the parameters affecting the structure of the haplotype cluster that is used to initialize the underlying Hidden Markov Model of BEAGLE. The number of markers with extremely high error rates for the maize datasets were more than halved by the use of a flint reference genome (F7, PE0075 etc.) instead of the commonly used B73. On average, error rates for imputation of ungenotyped markers were reduced by 8.5% by excluding genetically distant individuals from the reference panel for the chicken diversity panel. To optimize imputation accuracy one has to find a balance between representing as much of the genetic diversity as possible while avoiding the introduction of noise by including genetically distant individuals.

@article{Santos2020Noncrossing,
 author = {Santos, Bruno and Kneib, Thomas},
 year = {2020},
 title = {Noncrossing structured additive multiple-output Bayesian quantile regression models},
 pages = {825},
 volume = {97},
 number = {3},
 issn = {0960-3174},
 journal = {Statistics and Computing},
 doi = {10.1007/s11222-020-09925-x},
 howpublished = {refereed}
}

@article{Bee2019Estimating,
 author = {Bee, M. and Hambuckers, Julien and Trapin, L.},
 year = {2019},
 title = {Estimating Value-at-Risk for the g-and-h distribution: an indirect inference approach},
 keywords = {postdoc;stat},
 pages = {1255--1266},
 volume = {19},
 number = {8},
 issn = {1469-7688},
 journal = {Quantitative Finance},
 doi = {10.1080/14697688.2019.1580762},
 howpublished = {refereed}
}

@article{Fehrmann2019AUnified,
 author = {Fehrmann, Lutz and Kukunda, Collins B. and N{\"o}lke, Nils and Schnell, Sebastian and Seidel, Dominik and Magnussen, Steen and Kleinn, Christoph},
 year = {2019},
 title = {A unified framework for land cover monitoring based on a discrete global sampling grid {(GSG)}},
 url = {http://dx.doi.org/10.1007/s10661-018-7152-y},
 keywords = {ecol;phd},
 volume = {191},
 number = {2},
 issn = {0167-6369},
 journal = {Environmental Monitoring and Assessment},
 doi = {10.1007/s10661-018-7152-y},
 howpublished = {refereed}
}

@article{Freudenberg2019Large,
 author = {Freudenberg, Maximilian and N{\"o}lke, Nils and Agostini, Alejandro and Urban, Kira and W{\"o}rg{\"o}tter, Florentin and Kleinn, Christoph},
 year = {2019},
 title = {Large Scale Palm Tree Detection In High Resolution Satellite Images Using U-Net},
 keywords = {ecol;phd},
 pages = {312},
 volume = {11},
 number = {3},
 journal = {Remote Sensing},
 doi = {10.3390/rs11030312},
 howpublished = {refereed}
}

@article{Frioni2019Phasing,
 abstract = {BACKGROUND

Haplotype data contains more information than genotype data and provides possibilities such as imputing low frequency variants, inferring points of recombination, detecting recurrent mutations, mapping linkage disequilibrium (LD), studying selection signatures, estimating IBD probabilities, etc. In addition, haplotype structure is used to assess genetic diversity and expected accuracy in genomic selection programs. Nevertheless, the quality and efficiency of phasing has rarely been a subject of thorough study but was assessed mainly as a by-product in imputation quality studies. Moreover, phasing studies based on data of a poultry population are non-existent. The aim of this study was to evaluate the phasing quality of FImpute and Beagle, two of the most used phasing software.

RESULTS

We simulated ten replicated samples of a layer population comprising 888 individuals from a real SNP dataset of 580 k and a pedigree of 12 generations. Chromosomes analyzed were 1, 7 and 20. We measured the percentage of SNPs that were phased equally between true and phased haplotypes (Eqp), proportion of individuals completely correctly phased, number of incorrectly phased SNPs or Breakpoints (Bkp) and the length of inverted haplotype segments. Results were obtained for three different groups of individuals, with no parents or offspring genotyped in the dataset, with only one parent, and with both parents, respectively. The phasing was performed with Beagle (v3.3 and v4.1) and FImpute v2.2 (with and without pedigree). Eqp values ranged from 88 to 100{\%}, with the best results from haplotypes phased with Beagle v4.1 and FImpute with pedigree information and at least one parent genotyped. FImpute haplotypes showed a higher number of Bkp than Beagle. As a consequence, switched haplotype segments were longer for Beagle than for FImpute.

CONCLUSION

We concluded that for the dataset applied in this study Beagle v4.1 or FImpute with pedigree information and at least one parent genotyped in the data set were the best alternatives for obtaining high quality phased haplotypes.},
 author = {Frioni, N. and Cavero, D. and Simianer, H. and Erbe, Malena},
 year = {2019},
 title = {Phasing quality assessment in a brown layer population through family- and population-based software},
 keywords = {gen;phd},
 pages = {57},
 volume = {20},
 number = {1},
 issn = {1471-2156},
 journal = {BMC Genetics},
 doi = {10.1186/s12863-019-0759-3},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/31311514},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6636125},
 howpublished = {refereed}
}

BACKGROUND Haplotype data contains more information than genotype data and provides possibilities such as imputing low frequency variants, inferring points of recombination, detecting recurrent mutations, mapping linkage disequilibrium (LD), studying selection signatures, estimating IBD probabilities, etc. In addition, haplotype structure is used to assess genetic diversity and expected accuracy in genomic selection programs. Nevertheless, the quality and efficiency of phasing has rarely been a subject of thorough study but was assessed mainly as a by-product in imputation quality studies. Moreover, phasing studies based on data of a poultry population are non-existent. The aim of this study was to evaluate the phasing quality of FImpute and Beagle, two of the most used phasing software. RESULTS We simulated ten replicated samples of a layer population comprising 888 individuals from a real SNP dataset of 580 k and a pedigree of 12 generations. Chromosomes analyzed were 1, 7 and 20. We measured the percentage of SNPs that were phased equally between true and phased haplotypes (Eqp), proportion of individuals completely correctly phased, number of incorrectly phased SNPs or Breakpoints (Bkp) and the length of inverted haplotype segments. Results were obtained for three different groups of individuals, with no parents or offspring genotyped in the dataset, with only one parent, and with both parents, respectively. The phasing was performed with Beagle (v3.3 and v4.1) and FImpute v2.2 (with and without pedigree). Eqp values ranged from 88 to 100%, with the best results from haplotypes phased with Beagle v4.1 and FImpute with pedigree information and at least one parent genotyped. FImpute haplotypes showed a higher number of Bkp than Beagle. As a consequence, switched haplotype segments were longer for Beagle than for FImpute. CONCLUSION We concluded that for the dataset applied in this study Beagle v4.1 or FImpute with pedigree information and at least one parent genotyped in the data set were the best alternatives for obtaining high quality phased haplotypes.

@article{Graf2019Analyzing,
 author = {Graf, Wanda and Kleinn, Christoph and Schall, Peter and Nauss, Thomas and Detsch, Florian and Magdon, Paul},
 year = {2019},
 title = {Analyzing the relationship between historic canopy dynamics and current plant species diversity in the herb layer of temperate forests using long-term {L}andsat time series},
 keywords = {phd},
 pages = {111305},
 volume = {232},
 issn = {00344257},
 journal = {Remote Sensing of Environment},
 doi = {10.1016/j.rse.2019.111305},
 howpublished = {refereed}
}

@article{Grass2019Land,
 author = {Grass, Ingo and {[...]} and Libr{\'a}n-Embid, Felipe and Klaus, Felix and Udy, Kristy and Tscharntke, Teja},
 year = {2019},
 title = {Land--sharing/--sparing connectivity landscapes for ecosystem services and biodiversity conservation},
 url = {https://besjournals.onlinelibrary.wiley.com/doi/pdf/10.1002/pan3.21},
 pages = {262--272},
 volume = {1},
 number = {2},
 issn = {2575-8314},
 journal = {People and Nature},
 doi = {10.1002/pan3.21},
 file = {https://besjournals.onlinelibrary.wiley.com/doi/abs/10.1002/pan3.21},
 file = {https://besjournals.onlinelibrary.wiley.com/doi/full/10.1002/pan3.21},
 howpublished = {refereed}
}

@article{Groll2019LASSO,
 author = {Groll, Andreas and Hambuckers, Julien and Kneib, Thomas and Umlauf, Nikolaus},
 year = {2019},
 title = {{LASSO}-type penalization in the framework of generalized additive models for location, scale and shape},
 keywords = {postdoc;stat},
 pages = {59--73},
 volume = {140},
 issn = {01679473},
 journal = {Computational Statistics {\&} Data Analysis},
 doi = {10.1016/j.csda.2019.06.005},
 howpublished = {refereed}
}

@article{Haschka2019Provision,
 abstract = {The German health care system is among the most patient-oriented systems in Europe. Nevertheless, distinct utilisation patterns, access barriers due to socio-economic profiles, and potentials of misallocation of medical resources lead to disparities in the provision of health care services. We analyse how a possible over- and undersupply of services and the utilisation of and the access to the health care system relate to regional variations in the population's well-being. For this purpose, we employ a recent Bayesian stochastic frontier approach that allows for spatial dependence structures. Our results indicate that patient migration plays an important role in contributing to regional differences in the utilisation of the medical infrastructure. As a consequence, policy should take spatial patterns of health care utilisation into account to improve the allocation of medical resources.},
 author = {Haschka, Rouven Edgar and Schley, Katharina and Herwartz, Helmut},
 year = {2019},
 title = {Provision of health care services and regional diversity in {G}ermany: insights from a {B}ayesian health frontier analysis with spatial dependencies},
 keywords = {econ;phd},
 journal = {The European Journal of Health Economics},
 doi = {10.1007/s10198-019-01111-9.},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/31493180},
 howpublished = {refereed}
}

The German health care system is among the most patient-oriented systems in Europe. Nevertheless, distinct utilisation patterns, access barriers due to socio-economic profiles, and potentials of misallocation of medical resources lead to disparities in the provision of health care services. We analyse how a possible over- and undersupply of services and the utilisation of and the access to the health care system relate to regional variations in the population's well-being. For this purpose, we employ a recent Bayesian stochastic frontier approach that allows for spatial dependence structures. Our results indicate that patient migration plays an important role in contributing to regional differences in the utilisation of the medical infrastructure. As a consequence, policy should take spatial patterns of health care utilisation into account to improve the allocation of medical resources.

@article{Hesselbarth2019landscapemetrics,
 author = {Hesselbarth, Maximilian H. K. and Sciaini, Marco and With, Kimberly A. and Wiegand, Kerstin and Nowosad, Jakub},
 year = {2019},
 title = {{l}andscapemetrics: an open--source {R} tool to calculate landscape metrics},
 pages = {1648--1657},
 volume = {42},
 number = {10},
 issn = {09067590},
 journal = {Ecography},
 doi = {10.1111/ecog.04617},
 howpublished = {refereed}
}

@article{Klein2016Candidate,
 author = {Klein, Nadja and Entwistle, Andrew and Rosenberger, Albert and Kneib, Thomas and Bickeb{\"o}ller, Heike},
 year = {2019},
 title = {Candidate-gene association analysis for a continuous phenotype with a spike at zero using parent-offspring trios},
 keywords = {gen;phd;stat},
 pages = {1--15},
 volume = {60},
 issn = {0266-4763},
 journal = {Journal of Applied Statistics},
 doi = {10.1080/02664763.2019.1704226},
 howpublished = {refereed}
}

@article{Kukunda2019Scale,
 abstract = {Heterogeneity in forest structure, naturally occurring or induced by management, is continuous in space and time. However, measures used to quantify structure of forests are scale-variant, as they rely on bounded observations on either ecological or forest inventory observation units. The understanding of the influence of the scale of observation in mapping of forest structural heterogeneity is limited. Therefore, we researched into effects of plot size on quantifying forest structural heterogeneity, where we describe heterogeneity by three indices in stands under different management systems. In addition, we studied the performance of structural indices in separating different forest management systems across plot sizes, and created wall to wall maps of the indices using airborne LiDAR metrics describing the vertical distribution of canopy heights at different scales of observation. The studied indices are: Gini Coefficient (GC), Structural Complexity Index (SCI), and Enhanced Structural Complexity Index (ESCI). SCI and ESCI require fully mapped plots whereas GC has no information on individual tree locations. Inventory data from 95 one-hectare plots covering a range of management intensities from un-managed to age class forests were used. We quantified the three structural indices for 18 plot sizes ranging from 225 m2 to 10,000 m2. Linear fixed effects models were used to study the effects of plot sizes in different levels of structural heterogeneity and Random Forest (RF) models used to provide wall-to-wall maps at varying scales from airborne LiDAR data. The simulation showed that all indices were influenced by the scale of observation with larger effects for plots in forests with higher structural heterogeneity. For the data analyzed we found a threshold scale for enumerating stand structural heterogeneity between 900 m2 and 2500 m2. However, stable field and remote sensing predictions of stand structural heterogeneity required plots at least ⩾2500 m2. Compared to GC, SCI and ESCI improved separation of forest structure in the three management systems and at all observed scales. A change of plot sizes affected bivariate relationships between structural indices and airborne LiDAR metrics as well as the resultant predictive models. Smaller plot sizes yielded weaker relationships and predictive models. All structure indices were predicted from airborne LiDAR with RMSE⩽22{\%} at scales equal or larger than the identified threshold plot size. These findings are relevant to optimize plot sizes for efficient inventory and mapping of forest structural heterogeneity, as well as for the design of natural resource inventories. Additionally, derived maps are useful for studies on forest structure and the link with forest growth, degradation, management intensity, productivity, and biodiversity in the regions.},
 author = {Kukunda, Collins B. and Becksch{\"a}fer, Philip and Magdon, Paul and Schall, Peter and Wirth, Christian and Kleinn, Christoph},
 year = {2019},
 title = {Scale-guided mapping of forest stand structural heterogeneity from airborne {LiDAR}},
 keywords = {ecol;phd},
 pages = {410--425},
 volume = {102},
 issn = {1470160X},
 journal = {Ecological Indicators},
 doi = {10.1016/j.ecolind.2019.02.056},
 howpublished = {refereed}
}

Heterogeneity in forest structure, naturally occurring or induced by management, is continuous in space and time. However, measures used to quantify structure of forests are scale-variant, as they rely on bounded observations on either ecological or forest inventory observation units. The understanding of the influence of the scale of observation in mapping of forest structural heterogeneity is limited. Therefore, we researched into effects of plot size on quantifying forest structural heterogeneity, where we describe heterogeneity by three indices in stands under different management systems. In addition, we studied the performance of structural indices in separating different forest management systems across plot sizes, and created wall to wall maps of the indices using airborne LiDAR metrics describing the vertical distribution of canopy heights at different scales of observation. The studied indices are: Gini Coefficient (GC), Structural Complexity Index (SCI), and Enhanced Structural Complexity Index (ESCI). SCI and ESCI require fully mapped plots whereas GC has no information on individual tree locations. Inventory data from 95 one-hectare plots covering a range of management intensities from un-managed to age class forests were used. We quantified the three structural indices for 18 plot sizes ranging from 225 m2 to 10,000 m2. Linear fixed effects models were used to study the effects of plot sizes in different levels of structural heterogeneity and Random Forest (RF) models used to provide wall-to-wall maps at varying scales from airborne LiDAR data. The simulation showed that all indices were influenced by the scale of observation with larger effects for plots in forests with higher structural heterogeneity. For the data analyzed we found a threshold scale for enumerating stand structural heterogeneity between 900 m2 and 2500 m2. However, stable field and remote sensing predictions of stand structural heterogeneity required plots at least ⩾2500 m2. Compared to GC, SCI and ESCI improved separation of forest structure in the three management systems and at all observed scales. A change of plot sizes affected bivariate relationships between structural indices and airborne LiDAR metrics as well as the resultant predictive models. Smaller plot sizes yielded weaker relationships and predictive models. All structure indices were predicted from airborne LiDAR with RMSE⩽22% at scales equal or larger than the identified threshold plot size. These findings are relevant to optimize plot sizes for efficient inventory and mapping of forest structural heterogeneity, as well as for the design of natural resource inventories. Additionally, derived maps are useful for studies on forest structure and the link with forest growth, degradation, management intensity, productivity, and biodiversity in the regions.

@article{Li2019Integrating,
 abstract = {Gene expression profiles potentially hold valuable information for the prediction of breeding values and phenotypes. In this study, the utility of transcriptome data for phenotype prediction was tested with 185 inbred lines of Drosophila melanogaster for nine traits in two sexes. We incorporated the transcriptome data into genomic prediction via two methods: GTBLUP and GRBLUP, both combining single nucleotide polymorphisms (SNPs) and transcriptome data. The genotypic data was used to construct the common additive genomic relationship, which was used in genomic best linear unbiased prediction (GBLUP) or jointly in a linear mixed model with a transcriptome-based linear kernel (GTBLUP), or with a transcriptome-based Gaussian kernel (GRBLUP). We studied the predictive ability of the models and discuss a concept of {\textquotedbl}omics-augmented broad sense heritability{\textquotedbl} for the multi-omics era. For most traits, GRBLUP and GBLUP provided similar predictive abilities, but GRBLUP explained more of the phenotypic variance. There was only one trait (olfactory perception to Ethyl Butyrate in females) in which the predictive ability of GRBLUP (0.23) was significantly higher than the predictive ability of GBLUP (0.21). Our results suggest that accounting for transcriptome data has the potential to improve genomic predictions if transcriptome data can be included on a larger scale.},
 author = {Li, Zhengcao and Gao, Ning and Martini, Johannes W. R. and Simianer, Henner},
 year = {2019},
 title = {Integrating Gene Expression Data Into Genomic Prediction},
 keywords = {gen;phd},
 pages = {126},
 volume = {10},
 issn = {1664-8021},
 journal = {Frontiers in Genetics},
 doi = {10.3389/fgene.2019.00126},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/30858865},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6397893},
 howpublished = {refereed}
}

Gene expression profiles potentially hold valuable information for the prediction of breeding values and phenotypes. In this study, the utility of transcriptome data for phenotype prediction was tested with 185 inbred lines of Drosophila melanogaster for nine traits in two sexes. We incorporated the transcriptome data into genomic prediction via two methods: GTBLUP and GRBLUP, both combining single nucleotide polymorphisms (SNPs) and transcriptome data. The genotypic data was used to construct the common additive genomic relationship, which was used in genomic best linear unbiased prediction (GBLUP) or jointly in a linear mixed model with a transcriptome-based linear kernel (GTBLUP), or with a transcriptome-based Gaussian kernel (GRBLUP). We studied the predictive ability of the models and discuss a concept of \textquotedblomics-augmented broad sense heritability\textquotedbl for the multi-omics era. For most traits, GRBLUP and GBLUP provided similar predictive abilities, but GRBLUP explained more of the phenotypic variance. There was only one trait (olfactory perception to Ethyl Butyrate in females) in which the predictive ability of GRBLUP (0.23) was significantly higher than the predictive ability of GBLUP (0.21). Our results suggest that accounting for transcriptome data has the potential to improve genomic predictions if transcriptome data can be included on a larger scale.

@article{Martini2019Lost,
 abstract = {Mixed models can be considered as a type of penalized regression and are everyday tools in statistical genetics. The standard mixed model for whole genome regression (WGR) is ridge regression best linear unbiased prediction (RRBLUP) which is based on an additive marker effect model. Many publications have extended the additive WGR approach by incorporating interactions between loci or between genes and environment. In this context of penalized regressions with interactions, it has been reported that translating the coding of single nucleotide polymorphisms -for instance from -1,0,1 to 0,1,2- has an impact on the prediction of genetic values and interaction effects. In this work, we identify the reason for the relevance of variable coding in the general context of penalized polynomial regression. We show that in many cases, predictions of the genetic values are not invariant to translations of the variable coding, with an exception when only the sizes of the coefficients of monomials of highest total degree are penalized. The invariance of RRBLUP can be considered as a special case of this setting, with a polynomial of total degree 1, penalizing additive effects (total degree 1) but not the fixed effect (total degree 0). The extended RRBLUP (eRRBLUP), which includes interactions, is not invariant to translations because it does not only penalize interactions (total degree 2), but also additive effects (total degree 1). This observation implies that translation-invariance can be maintained in a pair-wise epistatic WGR if only interaction effects are penalized, but not the additive effects. In this regard, approaches of pre-selecting loci may not only reduce computation time, but can also help to avoid the variable coding issue. To illustrate the practical relevance, we compare different regressions on a publicly available wheat data set. We show that for an eRRBLUP, the relevance of the marker coding for interaction effect estimates increases with the number of variables included in the model. A biological interpretation of estimated interaction effects may therefore become more difficult. Consequently, comparing reproducing kernel Hilbert space (RKHS) approaches to WGR approaches modeling effects explicitly, the supposed advantage of an increased interpretability of the latter may not be real. Our theoretical results are generally valid for penalized regressions, for instance also for the least absolute shrinkage and selection operator (LASSO). Moreover, they apply to any type of interaction modeled by products of predictor variables in a penalized regression approach or by Hadamard products of covariance matrices in a mixed model.},
 author = {Martini, Johannes W. R. and Rosales, Francisco and Ha, Ngoc-Thuy and Heise, Johannes and Wimmer, Valentin and Kneib, Thomas},
 year = {2019},
 title = {Lost in Translation: {O}n the Problem of Data Coding in Penalized Whole Genome Regression with Interactions},
 keywords = {gen;phd},
 pages = {1117--1129},
 volume = {9},
 number = {4},
 journal = {G3: Genes|Genomes|Genetics},
 doi = {10.1534/g3.118.200961},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/30760541},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6469405},
 howpublished = {refereed}
}

Mixed models can be considered as a type of penalized regression and are everyday tools in statistical genetics. The standard mixed model for whole genome regression (WGR) is ridge regression best linear unbiased prediction (RRBLUP) which is based on an additive marker effect model. Many publications have extended the additive WGR approach by incorporating interactions between loci or between genes and environment. In this context of penalized regressions with interactions, it has been reported that translating the coding of single nucleotide polymorphisms -for instance from -1,0,1 to 0,1,2- has an impact on the prediction of genetic values and interaction effects. In this work, we identify the reason for the relevance of variable coding in the general context of penalized polynomial regression. We show that in many cases, predictions of the genetic values are not invariant to translations of the variable coding, with an exception when only the sizes of the coefficients of monomials of highest total degree are penalized. The invariance of RRBLUP can be considered as a special case of this setting, with a polynomial of total degree 1, penalizing additive effects (total degree 1) but not the fixed effect (total degree 0). The extended RRBLUP (eRRBLUP), which includes interactions, is not invariant to translations because it does not only penalize interactions (total degree 2), but also additive effects (total degree 1). This observation implies that translation-invariance can be maintained in a pair-wise epistatic WGR if only interaction effects are penalized, but not the additive effects. In this regard, approaches of pre-selecting loci may not only reduce computation time, but can also help to avoid the variable coding issue. To illustrate the practical relevance, we compare different regressions on a publicly available wheat data set. We show that for an eRRBLUP, the relevance of the marker coding for interaction effect estimates increases with the number of variables included in the model. A biological interpretation of estimated interaction effects may therefore become more difficult. Consequently, comparing reproducing kernel Hilbert space (RKHS) approaches to WGR approaches modeling effects explicitly, the supposed advantage of an increased interpretability of the latter may not be real. Our theoretical results are generally valid for penalized regressions, for instance also for the least absolute shrinkage and selection operator (LASSO). Moreover, they apply to any type of interaction modeled by products of predictor variables in a penalized regression approach or by Hadamard products of covariance matrices in a mixed model.

@article{Metzger2019Why,
 abstract = {Brazil's environmental legislation obliges private properties to retain a fixed proportion of their total area with native vegetation, the so-called ``Legal Reserves''. Those areas represent practically one third of the country's native vegetation and are well known for their role in biodiversity protection and in the provisioning of a wide range of ecosystem services for landowners and society. Despite their relevance, this instrument has been criticized by part of the agribusiness sector and its representatives in the Brazilian Congress. The Legal Reserve requirement is said to be too restrictive and to impede the full expansion of agricultural activities, and thus to be detrimental for the development of the country. Here, we critically analyze the arguments employed in the justification of a recently proposed bill that aims to completely extinguish Legal Reserves. We demonstrate that the arguments used are mostly unsupported by data, evidence or theory, besides being based on illogical reasoning. Further, we synthesize the principal benefits of Legal Reserves, including health and economic benefits, and emphasize the importance of these reserves for water, energy, food, and climate securities, in addition to their primary function of assisting in the maintenance of biodiversity in agricultural landscapes. We also highlight that Legal Reserves are a key-component for effective and less expensive nature-based solutions, and thus should be considered as assets for the development of Brazil rather than liabilities. Based on available sound scientific evidence and agreement on their relevance, we strongly oppose any attempt to extinguish or weaken the maintenance of Brazil's Legal Reserves.},
 author = {Metzger, Jean Paul and Bustamante, Mercedes M.C. and Ferreira, Joice and Fernandes, Geraldo Wilson and Libr{\'a}n-Embid, Felipe and Pillar, Val{\'e}rio D. and Prist, Paula R. and Rodrigues, Ricardo Ribeiro and Vieira, Ima C{\'e}lia G. and Overbeck, Gerhard E.},
 year = {2019},
 title = {Why {B}razil needs its Legal Reserves},
 url = {http://www.sciencedirect.com/science/article/pii/S253006441930118X},
 pages = {91--103},
 volume = {17},
 number = {3},
 issn = {2530-0644},
 journal = {Perspectives in Ecology and Conservation},
 doi = {10.1016/j.pecon.2019.07.002},
 howpublished = {refereed}
}

Brazil's environmental legislation obliges private properties to retain a fixed proportion of their total area with native vegetation, the so-called ``Legal Reserves''. Those areas represent practically one third of the country's native vegetation and are well known for their role in biodiversity protection and in the provisioning of a wide range of ecosystem services for landowners and society. Despite their relevance, this instrument has been criticized by part of the agribusiness sector and its representatives in the Brazilian Congress. The Legal Reserve requirement is said to be too restrictive and to impede the full expansion of agricultural activities, and thus to be detrimental for the development of the country. Here, we critically analyze the arguments employed in the justification of a recently proposed bill that aims to completely extinguish Legal Reserves. We demonstrate that the arguments used are mostly unsupported by data, evidence or theory, besides being based on illogical reasoning. Further, we synthesize the principal benefits of Legal Reserves, including health and economic benefits, and emphasize the importance of these reserves for water, energy, food, and climate securities, in addition to their primary function of assisting in the maintenance of biodiversity in agricultural landscapes. We also highlight that Legal Reserves are a key-component for effective and less expensive nature-based solutions, and thus should be considered as assets for the development of Brazil rather than liabilities. Based on available sound scientific evidence and agreement on their relevance, we strongly oppose any attempt to extinguish or weaken the maintenance of Brazil's Legal Reserves.

@article{Pook2019HaploBlocker,
 abstract = {The concept of haplotype blocks has been shown to be useful in genetics. Fields of application range from the detection of regions under positive selection to statistical methods that make use of dimension reduction. We propose a novel approach ({\textquotedbl}HaploBlocker{\textquotedbl}) for defining and inferring haplotype blocks that focuses on linkage instead of the commonly used population-wide measures of linkage disequilibrium. We define a haplotype block as a sequence of genetic markers that has a predefined minimum frequency in the population, and only haplotypes with a similar sequence of markers are considered to carry that block, effectively screening a dataset for group-wise identity-by-descent. From these haplotype blocks, we construct a haplotype library that represents a large proportion of genetic variability with a limited number of blocks. Our method is implemented in the associated R-package HaploBlocker, and provides flexibility not only to optimize the structure of the obtained haplotype library for subsequent analyses, but also to handle datasets of different marker density and genetic diversity. By using haplotype blocks instead of single nucleotide polymorphisms (SNPs), local epistatic interactions can be naturally modeled, and the reduced number of parameters enables a wide variety of new methods for further genomic analyses such as genomic prediction and the detection of selection signatures. We illustrate our methodology with a dataset comprising 501 doubled haploid lines in a European maize landrace genotyped at 501,124 SNPs. With the suggested approach, we identified 2991 haplotype blocks with an average length of 2685 SNPs that together represent 94{\%} of the dataset.},
 author = {Pook, Torsten and Schlather, Martin and de {los Campos}, Gustavo and Mayer, Manfred and Schoen, Chris Carolin and Simianer, Henner},
 year = {2019},
 title = {{HaploBlocker}: Creation of Subgroup-Specific Haplotype Blocks and Libraries},
 keywords = {gen;phd},
 pages = {1045--1061},
 volume = {212},
 number = {4},
 issn = {1943-2631},
 journal = {Genetics},
 doi = {10.1534/genetics.119.302283},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/31152070},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6707469},
 howpublished = {refereed}
}

The concept of haplotype blocks has been shown to be useful in genetics. Fields of application range from the detection of regions under positive selection to statistical methods that make use of dimension reduction. We propose a novel approach (\textquotedblHaploBlocker\textquotedbl) for defining and inferring haplotype blocks that focuses on linkage instead of the commonly used population-wide measures of linkage disequilibrium. We define a haplotype block as a sequence of genetic markers that has a predefined minimum frequency in the population, and only haplotypes with a similar sequence of markers are considered to carry that block, effectively screening a dataset for group-wise identity-by-descent. From these haplotype blocks, we construct a haplotype library that represents a large proportion of genetic variability with a limited number of blocks. Our method is implemented in the associated R-package HaploBlocker, and provides flexibility not only to optimize the structure of the obtained haplotype library for subsequent analyses, but also to handle datasets of different marker density and genetic diversity. By using haplotype blocks instead of single nucleotide polymorphisms (SNPs), local epistatic interactions can be naturally modeled, and the reduced number of parameters enables a wide variety of new methods for further genomic analyses such as genomic prediction and the detection of selection signatures. We illustrate our methodology with a dataset comprising 501 doubled haploid lines in a European maize landrace genotyped at 501,124 SNPs. With the suggested approach, we identified 2991 haplotype blocks with an average length of 2685 SNPs that together represent 94% of the dataset.

@article{Thaden2017General,
 abstract = {Modeling complex relationships and interactions between variables is an

ongoing statistical challenge. In particular, the joint modeling of multiple

response variables has recently gained interest among methodological and

applied researchers. In this article, we contribute to this development by

incorporating semiparametric predictors into recursive simultaneous equation

models. In particular, we extend the existing framework by imposing effect

priors that account for correlation of the effects across equations. This idea

can be seen as a generalization of multivariate conditional autoregressive

priors used for the analysis of multivariate spatial data.

We implement a Gibbs sampler for the estimation and evaluate the model in

an elaborate simulation study. Finally, we illustrate the applicability of our

approach with real data examples on malnutrition in Asia and Africa as well

as the analysis of plant and species richness with respect to environmental

diversity.},
 author = {Thaden, Hauke and Klein, Nadja and Kneib, Thomas},
 year = {2019},
 title = {Multivariate effect priors in bivariate semiparametric recursive Gaussian models},
 url = {https://www.uni-goettingen.de/de/13_Thaden_02_2017/558175.html},
 keywords = {econ;phd},
 pages = {51--66},
 volume = {137},
 issn = {01679473},
 journal = {Computational Statistics {\&} Data Analysis},
 doi = {10.1016/j.csda.2018.12.004},
 howpublished = {refereed}
}

Modeling complex relationships and interactions between variables is an ongoing statistical challenge. In particular, the joint modeling of multiple response variables has recently gained interest among methodological and applied researchers. In this article, we contribute to this development by incorporating semiparametric predictors into recursive simultaneous equation models. In particular, we extend the existing framework by imposing effect priors that account for correlation of the effects across equations. This idea can be seen as a generalization of multivariate conditional autoregressive priors used for the analysis of multivariate spatial data. We implement a Gibbs sampler for the estimation and evaluate the model in an elaborate simulation study. Finally, we illustrate the applicability of our approach with real data examples on malnutrition in Asia and Africa as well as the analysis of plant and species richness with respect to environmental diversity.

@article{Tozzi2019Global,
 abstract = {Genome-wide association studies have led in the past to the discovery of susceptibility genes for many diseases including cancer and inflammatory conditions. However, a number of these studies did not realise their full potential. A first critical step in developing such large-scale studies is the choice of genotyping array with respect to the study goal. Coverage is the central criterion for array evaluation. We distinguish between estimates of global coverage across the genome, coverage for each chromosome, coverage for selected pathways and the coverage for genes of interest. Here, we focus on inflammatory and immunological pathways and genes relevant for haematopoietic stem cell transplantation. We compared three arrays: the Infinium Global Screening Array-24 v1.0, the Infinium OncoArray-500 K BeadChip and the Infinium PsychArray-24 v1.2 BeadChip. We employed the European population from the 1000 Genomes Project as reference genome. Global coverage was found to range between 12.2 and 14.2{\%} whereas coverage for a selected pathway ranged from 6.2 to 13.2{\%} and gene coverage ranged from 0 to 54.1{\%}. The Global Screening Array outperformed both other arrays in terms of global coverage, for most chromosomes, most considered pathways and most genes. When selecting suitable arrays for a new study, the coverage of pathways or genes of interest should be considered in addition to global coverage. Local coverage should be regarded when discussing association findings inconsistent across studies and can be useful in data analysis and decision making for additional genotyping.},
 author = {Tozzi, Viola and Rosenberger, Albert and Kube, Dieter and Bickeb{\"o}ller, Heike},
 year = {2019},
 title = {Global, pathway and gene coverage of three Illumina arrays with respect to inflammatory and immune-related pathways},
 pages = {1716--1723},
 volume = {27},
 number = {11},
 journal = {European Journal of Human Genetics},
 doi = {10.1038/s41431-019-0441-2},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/31227809},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6871524},
 howpublished = {refereed}
}

Genome-wide association studies have led in the past to the discovery of susceptibility genes for many diseases including cancer and inflammatory conditions. However, a number of these studies did not realise their full potential. A first critical step in developing such large-scale studies is the choice of genotyping array with respect to the study goal. Coverage is the central criterion for array evaluation. We distinguish between estimates of global coverage across the genome, coverage for each chromosome, coverage for selected pathways and the coverage for genes of interest. Here, we focus on inflammatory and immunological pathways and genes relevant for haematopoietic stem cell transplantation. We compared three arrays: the Infinium Global Screening Array-24 v1.0, the Infinium OncoArray-500 K BeadChip and the Infinium PsychArray-24 v1.2 BeadChip. We employed the European population from the 1000 Genomes Project as reference genome. Global coverage was found to range between 12.2 and 14.2% whereas coverage for a selected pathway ranged from 6.2 to 13.2% and gene coverage ranged from 0 to 54.1%. The Global Screening Array outperformed both other arrays in terms of global coverage, for most chromosomes, most considered pathways and most genes. When selecting suitable arrays for a new study, the coverage of pathways or genes of interest should be considered in addition to global coverage. Local coverage should be regarded when discussing association findings inconsistent across studies and can be useful in data analysis and decision making for additional genotyping.

@article{Vollmer2019Measuring,
 author = {Vollmer, Teresa and Herwartz, Helmut and von Cramon-Taubadel, Stephan},
 year = {2019},
 title = {Measuring price discovery in the {E}uropean wheat market using the partial cointegration approach},
 keywords = {PI},
 pages = {441},
 volume = {109},
 number = {2},
 issn = {0165-1587},
 journal = {European Review of Agricultural Economics},
 doi = {10.1093/erae/jbz040},
 howpublished = {refereed}
}

@article{Yasmeen2019Polygenic,
 author = {Yasmeen, Summaira and Papiol, Sergi and Falkai, Peter and Schulze, Thomas G. and Bickeb{\"o}ller, Heike},
 year = {2019},
 title = {Polygenic Risk for Schizophrenia and Global Assessment of Functioning -- A Comparison with In-Silico Data},
 keywords = {gen;phd},
 journal = {Journal of Psychiatry and Brain Science},
 doi = {10.20900/jpbs.20190003},
 howpublished = {refereed}
}

@article{Beduschi2018Spatial,
 author = {Beduschi, Tatiane and Kormann, Urs and Tscharntke, Teja and Scherber, Christoph},
 year = {2018},
 title = {Spatial community turnover of pollinators is relaxed by semi-natural habitats, but not by mass-flowering crops in agricultural landscapes},
 url = {http://dx.doi.org/10.1016/j.biocon.2018.01.016},
 keywords = {ecol;phd},
 pages = {59--66},
 volume = {221},
 issn = {00063207},
 journal = {Biological Conservation},
 doi = {10.1016/j.biocon.2018.01.016},
 howpublished = {refereed}
}

@article{Beissinger2018Simple,
 abstract = {Important traits in agricultural, natural, and human populations are increasingly being shown to be under the control of many genes that individually contribute only a small proportion of genetic variation. However, the majority of modern tools in quantitative and population genetics, including genome-wide association studies and selection-mapping protocols, are designed to identify individual genes with large effects. We have developed an approach to identify traits that have been under selection and are controlled by large numbers of loci. In contrast to existing methods, our technique uses additive-effects estimates from all available markers, and relates these estimates to allele-frequency change over time. Using this information, we generate a composite statistic, denoted [Formula: see text] which can be used to test for significant evidence of selection on a trait. Our test requires pre- and postselection genotypic data but only a single time point with phenotypic information. Simulations demonstrate that [Formula: see text] is powerful for identifying selection, particularly in situations where the trait being tested is controlled by many genes, which is precisely the scenario where classical approaches for selection mapping are least powerful. We apply this test to breeding populations of maize and chickens, where we demonstrate the successful identification of selection on traits that are documented to have been under selection.},
 author = {Beissinger, Tim and Kruppa, Jochen and Cavero, David and Ha, Ngoc-Thuy and Erbe, Malena and Simianer, Henner},
 year = {2018},
 title = {A Simple Test Identifies Selection on Complex Traits},
 keywords = {gen;postdoc},
 pages = {321--333},
 volume = {209},
 number = {1},
 issn = {1943-2631},
 journal = {Genetics},
 doi = {10.1534/genetics.118.300857},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/29545467},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5937188},
 howpublished = {refereed}
}

Important traits in agricultural, natural, and human populations are increasingly being shown to be under the control of many genes that individually contribute only a small proportion of genetic variation. However, the majority of modern tools in quantitative and population genetics, including genome-wide association studies and selection-mapping protocols, are designed to identify individual genes with large effects. We have developed an approach to identify traits that have been under selection and are controlled by large numbers of loci. In contrast to existing methods, our technique uses additive-effects estimates from all available markers, and relates these estimates to allele-frequency change over time. Using this information, we generate a composite statistic, denoted [Formula: see text] which can be used to test for significant evidence of selection on a trait. Our test requires pre- and postselection genotypic data but only a single time point with phenotypic information. Simulations demonstrate that [Formula: see text] is powerful for identifying selection, particularly in situations where the trait being tested is controlled by many genes, which is precisely the scenario where classical approaches for selection mapping are least powerful. We apply this test to breeding populations of maize and chickens, where we demonstrate the successful identification of selection on traits that are documented to have been under selection.

@article{Hambuckers2018Marcov-switching,
 author = {Hambuckers, Julien and Kneib, Thomas and Langrock, R. and Silbersdorff, A.},
 year = {2018},
 title = {A {M}arkov-switching generalized additive model for compound {P}oisson processes, with applications to operational loss models},
 url = {http://dx.doi.org/10.1080/14697688.2017.1417625},
 keywords = {postdoc;stat},
 pages = {1679--1698},
 volume = {18},
 number = {10},
 issn = {1469-7688},
 journal = {Quantitative Finance},
 doi = {10.1080/14697688.2017.1417625},
 howpublished = {refereed}
}

@article{Hambuckers2018Understanding,
 author = {Hambuckers, Julien and Groll, Andreas and Kneib, Thomas},
 year = {2018},
 title = {Understanding the economic determinants of the severity of operational losses: A regularized generalized {P}areto regression approach},
 url = {http://dx.doi.org/10.1002/jae.2638},
 keywords = {econ;postdoc;stat},
 issn = {08837252},
 journal = {Journal of Applied Econometrics},
 doi = {10.1002/jae.2638},
 howpublished = {refereed}
}

@article{Heise2018Phenotypic,
 abstract = {The aim of this study was to answer the question whether models for genetic evaluations of longevity should include a correction for age at first calving (AFC). For this purpose, phenotypic and genetic relationships between AFC, its component traits age at first insemination (AFI) and interval from first to last insemination (FLI), and survival of different periods of the first lactation (S1: 0 to 49 d, S2: 50 to 249 d, S3: 250 d to second calving) were investigated. Data of 721,919 German Holstein heifers, being inseminated for the first time during the years from 2003 to 2012, were used for the analyses. Phenotypic correlations of AFI, FLI, and AFC to S1 to S3 were negative. Mean estimated heritabilities were 0.239 (AFI), 0.007 (FLI), and 0.103 (AFC) and 0.023 (S1), 0.016 (S2), and 0.028 (S3) on the observed scale. The genetic correlation between AFI and FLI was close to zero. Genetic correlations between AFI and the survival traits were -0.08 (S1), -0.02 (S2), and -0.10 (S3); those between FLI and the survival traits were -0.14 (S1), -0.20 (S2), and -0.44 (S3); and those between AFC and the survival traits were -0.09 (S1), -0.06 (S2), and -0.20 (S3). Some of these genetic correlations were different from zero, which suggests that correcting for AFC in genetic evaluations for longevity in dairy cows might remove functional genetic variance and should be reconsidered.},
 author = {Heise, Johannes and Stock, Kathrin F. and Reinhardt, Friedrich and Ha, Ngoc-Thuy and Simianer, Henner},
 year = {2018},
 title = {Phenotypic and genetic relationships between age at first calving, its component traits, and survival of heifers up to second calving},
 keywords = {gen;postdoc},
 pages = {425--432},
 volume = {101},
 number = {1},
 journal = {Journal of Dairy Science},
 doi = {10.3168/jds.2017-12957},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/29128222},
 howpublished = {refereed}
}

The aim of this study was to answer the question whether models for genetic evaluations of longevity should include a correction for age at first calving (AFC). For this purpose, phenotypic and genetic relationships between AFC, its component traits age at first insemination (AFI) and interval from first to last insemination (FLI), and survival of different periods of the first lactation (S1: 0 to 49 d, S2: 50 to 249 d, S3: 250 d to second calving) were investigated. Data of 721,919 German Holstein heifers, being inseminated for the first time during the years from 2003 to 2012, were used for the analyses. Phenotypic correlations of AFI, FLI, and AFC to S1 to S3 were negative. Mean estimated heritabilities were 0.239 (AFI), 0.007 (FLI), and 0.103 (AFC) and 0.023 (S1), 0.016 (S2), and 0.028 (S3) on the observed scale. The genetic correlation between AFI and FLI was close to zero. Genetic correlations between AFI and the survival traits were -0.08 (S1), -0.02 (S2), and -0.10 (S3); those between FLI and the survival traits were -0.14 (S1), -0.20 (S2), and -0.44 (S3); and those between AFC and the survival traits were -0.09 (S1), -0.06 (S2), and -0.20 (S3). Some of these genetic correlations were different from zero, which suggests that correcting for AFC in genetic evaluations for longevity in dairy cows might remove functional genetic variance and should be reconsidered.

@article{Herwartz2018Improving,
 abstract = {The provision of health care in Germany exhibits sizeable geographic variation with a heterogeneous allocation of medical services in rural and urban areas. Furthermore, distinct utilisation patterns and access barriers due to the socio-economic environment might cause inefficiencies in the provision of health care services. Accordingly, an improved understanding of factors governing inefficiencies in health care provision is likely to benefit an efficient spatial allocation of health care infrastructure. We analyse how socio-economic factors influence the regional distribution of (in)efficiencies in the provision of health care services by means of a stochastic frontier analysis. Our results highlight that regional deprivation relates to inefficient provision of health care services. As a consequence, policies should also consider socio-economic conditions to improve the allocation of medical services and overall health.},
 author = {Herwartz, Helmut and Schley, Katharina},
 year = {2018},
 title = {Improving health care service provision by adapting to regional diversity: An efficiency analysis for the case of {G}ermany},
 keywords = {econ;phd},
 pages = {293--300},
 volume = {122},
 number = {3},
 issn = {0168-8510},
 journal = {Health Policy},
 doi = {10.1016/j.healthpol.2018.01.004},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/29482949},
 howpublished = {refereed}
}

The provision of health care in Germany exhibits sizeable geographic variation with a heterogeneous allocation of medical services in rural and urban areas. Furthermore, distinct utilisation patterns and access barriers due to the socio-economic environment might cause inefficiencies in the provision of health care services. Accordingly, an improved understanding of factors governing inefficiencies in health care provision is likely to benefit an efficient spatial allocation of health care infrastructure. We analyse how socio-economic factors influence the regional distribution of (in)efficiencies in the provision of health care services by means of a stochastic frontier analysis. Our results highlight that regional deprivation relates to inefficient provision of health care services. As a consequence, policies should also consider socio-economic conditions to improve the allocation of medical services and overall health.

@article{Hesselbarth2018Density,
 abstract = {Savannas can be defined by the co-dominance of grasses and trees. Interactions between these two life forms are relatively well studied, whereas tree-tree interactions attracted increased attention only recently. However, the influence of woody plant density on tree-tree interactions is rarely considered. We studied tree-tree interactions in a semi-arid and a mesic savanna to test for differences between open and dense woody vegetation in relation to broad-scale environmental conditions. We applied spatial point pattern analysis to gain a better understanding of processes, such as competition, facilitation and disturbances, affecting the spatial distribution of trees. Competition between trees was most pronounced in dense vegetation, whereas facilitation effects were more common in open vegetation. Further, we found that factors shaping the spatial patterns differ with scale. At short tree-to-tree distances, results indicate limited seed dispersal as the most influential factor explaining the spatial distribution of trees. However, with increasing tree-to-tree distances, environmental heterogeneity in the semi-arid savanna and disturbances in the mesic savanna became more important. We conclude that studying tree-tree interactions in savannas should explicitly consider the actual woody plant density, especially when different savanna types are compared.},
 author = {Hesselbarth, Maximilian H. K. and Wiegand, K. and Dreber, N. and Kellner, K. and Esser, Daniel S. and Tsvuura, Z.},
 year = {2018},
 title = {Density-dependent spatial patterning of woody plants differs between a semi-arid and a mesic savanna in {South Africa}},
 keywords = {ecol;phd},
 pages = {103--112},
 volume = {157},
 issn = {01401963},
 journal = {Journal of Arid Environments},
 doi = {10.1016/j.jaridenv.2018.06.002},
 howpublished = {refereed}
}

Savannas can be defined by the co-dominance of grasses and trees. Interactions between these two life forms are relatively well studied, whereas tree-tree interactions attracted increased attention only recently. However, the influence of woody plant density on tree-tree interactions is rarely considered. We studied tree-tree interactions in a semi-arid and a mesic savanna to test for differences between open and dense woody vegetation in relation to broad-scale environmental conditions. We applied spatial point pattern analysis to gain a better understanding of processes, such as competition, facilitation and disturbances, affecting the spatial distribution of trees. Competition between trees was most pronounced in dense vegetation, whereas facilitation effects were more common in open vegetation. Further, we found that factors shaping the spatial patterns differ with scale. At short tree-to-tree distances, results indicate limited seed dispersal as the most influential factor explaining the spatial distribution of trees. However, with increasing tree-to-tree distances, environmental heterogeneity in the semi-arid savanna and disturbances in the mesic savanna became more important. We conclude that studying tree-tree interactions in savannas should explicitly consider the actual woody plant density, especially when different savanna types are compared.

@article{Kubitza2018Land,
 author = {Kubitza, Christoph and Krishna, Vijesh V. and Urban, Kira and Alamsyah, Zulkifli and Qaim, Matin},
 year = {2018},
 title = {Land Property Rights, Agricultural Intensification, and Deforestation in {I}ndonesia},
 pages = {312--321},
 volume = {147},
 issn = {09218009},
 journal = {Ecological Economics},
 doi = {10.1016/j.ecolecon.2018.01.021},
 howpublished = {refereed}
}

@article{Kukunda2018Ensemble,
 author = {Kukunda, Collins B. and Duque-Lazo, Joaqu$\backslash$'{\i}n and Gonz{\'a}lez-Ferreiro, Eduardo and Thaden, Hauke and Kleinn, Christoph},
 year = {2018},
 title = {Ensemble classification of individual  \textit{Pinus}  crowns from multispectral satellite imagery and airborne {LiDAR}},
 url = {http://dx.doi.org/10.1016/j.jag.2017.09.016},
 keywords = {ecol;phd},
 pages = {12--23},
 volume = {65},
 issn = {03032434},
 journal = {International Journal of Applied Earth Observation and Geoinformation},
 doi = {10.1016/j.jag.2017.09.016},
 howpublished = {refereed}
}

@article{Martini2018TheEffect,
 abstract = {BACKGROUND

The single-step covariance matrix H combines the pedigree-based relationship matrix [Formula: see text] with the more accurate information on realized relatedness of genotyped individuals represented by the genomic relationship matrix [Formula: see text]. In particular, to improve convergence behavior of iterative approaches and to reduce inflation, two weights [Formula: see text] and [Formula: see text] have been introduced in the definition of [Formula: see text], which blend the inverse of a part of [Formula: see text] with the inverse of [Formula: see text]. Since the definition of this blending is based on the equation describing [Formula: see text], its impact on the structure of [Formula: see text] is not obvious. In a joint discussion, we considered the question of the shape of [Formula: see text] for non-trivial [Formula: see text] and [Formula: see text].

RESULTS

Here, we present the general matrix [Formula: see text] as a function of these parameters and discuss its structure and properties. Moreover, we screen for optimal values of [Formula: see text] and [Formula: see text] with respect to predictive ability, inflation and iterations up to convergence on a well investigated, publicly available wheat data set.

CONCLUSION

Our results may help the reader to develop a better understanding for the effects of changes of [Formula: see text] and [Formula: see text] on the covariance model. In particular, we give theoretical arguments that as a general tendency, inflation will be reduced by increasing [Formula: see text] or by decreasing [Formula: see text].},
 author = {Martini, Johannes W. R. and Schrauf, Matias F. and Garcia-Baccino, Carolina A. and Pimentel, Eduardo C. G. and {[...]} and Simianer, Henner},
 year = {2018},
 title = {The effect of the H\textsuperscript{-1} scaling factors \textgreek{t} and \textgreek{w} on the structure of H in the single-step procedure},
 keywords = {gen;phd},
 pages = {16},
 volume = {50},
 number = {1},
 journal = {Genetics, Selection, Evolution : GSE},
 doi = {10.1186/s12711-018-0386-x},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/29653506},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5899415},
 howpublished = {refereed}
}

BACKGROUND The single-step covariance matrix H combines the pedigree-based relationship matrix [Formula: see text] with the more accurate information on realized relatedness of genotyped individuals represented by the genomic relationship matrix [Formula: see text]. In particular, to improve convergence behavior of iterative approaches and to reduce inflation, two weights [Formula: see text] and [Formula: see text] have been introduced in the definition of [Formula: see text], which blend the inverse of a part of [Formula: see text] with the inverse of [Formula: see text]. Since the definition of this blending is based on the equation describing [Formula: see text], its impact on the structure of [Formula: see text] is not obvious. In a joint discussion, we considered the question of the shape of [Formula: see text] for non-trivial [Formula: see text] and [Formula: see text]. RESULTS Here, we present the general matrix [Formula: see text] as a function of these parameters and discuss its structure and properties. Moreover, we screen for optimal values of [Formula: see text] and [Formula: see text] with respect to predictive ability, inflation and iterations up to convergence on a well investigated, publicly available wheat data set. CONCLUSION Our results may help the reader to develop a better understanding for the effects of changes of [Formula: see text] and [Formula: see text] on the covariance model. In particular, we give theoretical arguments that as a general tendency, inflation will be reduced by increasing [Formula: see text] or by decreasing [Formula: see text].

@article{Radny2018Influence,
 author = {Radny, Janina and {van der Putten}, Wim H. and Tielb{\"o}rger, Katja and Meyer, Katrin M.},
 year = {2018},
 title = {Influence of seed size on performance of non-native annual plant species in a novel community at two planting densities},
 keywords = {ecol;phd},
 pages = {131--137},
 volume = {92},
 issn = {1146609X},
 journal = {Acta Oecologica},
 doi = {10.1016/j.actao.2018.05.005},
 howpublished = {refereed}
}

@article{Radny2018TheRole,
 abstract = {Background

Establishment success of non-native species is not only influenced by environmental conditions, but also by interactions with local competitors and enemies. The magnitude of these biotic interactions is mediated by species traits that reflect competitive strength or defence mechanisms. Our aim was to investigate the importance of species traits for successful establishment of non-native species in a native community exhibiting biotic resistance in the form of competition and herbivory.

Methods

We developed a trait-based, individual-based simulation model tracking the survival of non-native plants in a native community. In the model, non-native plants are characterized by high or low values of competition and defence traits. Model scenarios included variation of initial number of non-natives, intensity of competitive interaction, density of herbivores and density as well as mixture of the native community.

Results

Traits related to competition had a much greater impact on survival of non-native species than traits related to defence. Survival rates of strong competitors never fell below 50{\%} while survival of weak competitors averaged at about 10{\%}. Weak competitors were also much more susceptible to competitive pressures such as community density, composition and competition intensity. Strong competitors responded negatively to changes in competition intensity, but hardly to composition or density of the native community. High initial numbers of non-native individuals decreased survival rate of strong competitors, but increased the survival rate of weak competitors. Survival under herbivore attack was only slightly higher for plants with high defensive ability than for those with low defensive ability. Surprisingly, though, herbivory increased survival of species classified as weak competitors.

Discussion

High survival rates of strong non-native competitors relate to a higher probability of successful establishment than for weak competitors. However, the reduced survival of strong competitors at high initial numbers indicates a self-thinning effect, probably mediated by a strongly competitive milieu. For weak competitors, our model emphasizes positive effects of high propagule pressure known from field studies. General effects of herbivory or defence abilities on survival were not supported by our model. However, the positive effect of herbivory on survival of weak competitors indicated side effects of herbivory, such as weakening resident competitors. This might play an important role for establishment of non-natives in a new community.},
 author = {Radny, Janina and Meyer, Katrin M.},
 year = {2018},
 title = {The role of biotic factors during plant establishment in novel communities assessed with an agent-based simulation model},
 keywords = {ecol;phd},
 pages = {e5342},
 volume = {6},
 issn = {2167-8359},
 journal = {PeerJ},
 doi = {10.7717/peerj.5342},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/30123698},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6087422},
 howpublished = {refereed}
}

Background Establishment success of non-native species is not only influenced by environmental conditions, but also by interactions with local competitors and enemies. The magnitude of these biotic interactions is mediated by species traits that reflect competitive strength or defence mechanisms. Our aim was to investigate the importance of species traits for successful establishment of non-native species in a native community exhibiting biotic resistance in the form of competition and herbivory. Methods We developed a trait-based, individual-based simulation model tracking the survival of non-native plants in a native community. In the model, non-native plants are characterized by high or low values of competition and defence traits. Model scenarios included variation of initial number of non-natives, intensity of competitive interaction, density of herbivores and density as well as mixture of the native community. Results Traits related to competition had a much greater impact on survival of non-native species than traits related to defence. Survival rates of strong competitors never fell below 50% while survival of weak competitors averaged at about 10%. Weak competitors were also much more susceptible to competitive pressures such as community density, composition and competition intensity. Strong competitors responded negatively to changes in competition intensity, but hardly to composition or density of the native community. High initial numbers of non-native individuals decreased survival rate of strong competitors, but increased the survival rate of weak competitors. Survival under herbivore attack was only slightly higher for plants with high defensive ability than for those with low defensive ability. Surprisingly, though, herbivory increased survival of species classified as weak competitors. Discussion High survival rates of strong non-native competitors relate to a higher probability of successful establishment than for weak competitors. However, the reduced survival of strong competitors at high initial numbers indicates a self-thinning effect, probably mediated by a strongly competitive milieu. For weak competitors, our model emphasizes positive effects of high propagule pressure known from field studies. General effects of herbivory or defence abilities on survival were not supported by our model. However, the positive effect of herbivory on survival of weak competitors indicated side effects of herbivory, such as weakening resident competitors. This might play an important role for establishment of non-natives in a new community.

@article{Reimer2018Analysis,
 abstract = {BACKGROUND

Domestication has led to substantial phenotypic and genetic variation in domestic animals. In pigs, the size of so called minipigs differs by one order of magnitude compared to breeds of large body size. We used biallelic SNPs identified from re-sequencing data to compare various publicly available wild and domestic populations against two minipig breeds to gain better understanding of the genetic background of the extensive body size variation. We combined two complementary measures, expected heterozygosity and the composite likelihood ratio test implemented in {\textquotedbl}SweepFinder{\textquotedbl}, to identify signatures of selection in Minipigs. We intersected these sweep regions with a measure of differentiation, namely FST, to remove regions of low variation across pigs. An extraordinary large sweep between 52 and 61~Mb on chromosome X was separately analyzed based on SNP-array data of F2 individuals from a cross of Goettingen Minipigs and large pigs.

RESULTS

Selective sweep analysis identified putative sweep regions for growth and subsequent gene annotation provided a comprehensive set of putative candidate genes. A long swept haplotype on chromosome X, descending from the Goettingen Minipig founders was associated with a reduction of adult body length by 3{\%} in F2 cross-breds.

CONCLUSION

The resulting set of genes in putative sweep regions implies that the genetic background of body size variation in pigs is polygenic rather than mono- or oligogenic. Identified genes suggest alterations in metabolic functions and a possible insulin resistance to contribute to miniaturization. A size QTL located within the sweep on chromosome X, with an estimated effect of 3{\%} on body length, is comparable to the largest known in pigs or other species. The androgen receptor AR, previously known to influence pig performance and carcass traits, is the most obvious potential candidate gene within this region.},
 author = {Reimer, C. and Rubin, C-J and Sharifi, A. R. and Ha, Ngoc-Thuy and {[...]} and Schlather, M. and Simianer, H.},
 year = {2018},
 title = {Analysis of porcine body size variation using re-sequencing data of miniature and large pigs},
 keywords = {gen;postdoc},
 pages = {687},
 volume = {19},
 number = {1},
 journal = {BMC Genomics},
 doi = {10.1186/s12864-018-5009-y},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/30231878},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6146782},
 howpublished = {refereed}
}

BACKGROUND Domestication has led to substantial phenotypic and genetic variation in domestic animals. In pigs, the size of so called minipigs differs by one order of magnitude compared to breeds of large body size. We used biallelic SNPs identified from re-sequencing data to compare various publicly available wild and domestic populations against two minipig breeds to gain better understanding of the genetic background of the extensive body size variation. We combined two complementary measures, expected heterozygosity and the composite likelihood ratio test implemented in \textquotedblSweepFinder\textquotedbl, to identify signatures of selection in Minipigs. We intersected these sweep regions with a measure of differentiation, namely FST, to remove regions of low variation across pigs. An extraordinary large sweep between 52 and 61 Mb on chromosome X was separately analyzed based on SNP-array data of F2 individuals from a cross of Goettingen Minipigs and large pigs. RESULTS Selective sweep analysis identified putative sweep regions for growth and subsequent gene annotation provided a comprehensive set of putative candidate genes. A long swept haplotype on chromosome X, descending from the Goettingen Minipig founders was associated with a reduction of adult body length by 3% in F2 cross-breds. CONCLUSION The resulting set of genes in putative sweep regions implies that the genetic background of body size variation in pigs is polygenic rather than mono- or oligogenic. Identified genes suggest alterations in metabolic functions and a possible insulin resistance to contribute to miniaturization. A size QTL located within the sweep on chromosome X, with an estimated effect of 3% on body length, is comparable to the largest known in pigs or other species. The androgen receptor AR, previously known to influence pig performance and carcass traits, is the most obvious potential candidate gene within this region.

@article{Schley2018Health,
 author = {Schley, Katharina},
 year = {2018},
 title = {Health care service provision in {Europe} and regional diversity: a stochastic metafrontier approach},
 url = {http://dx.doi.org/10.1186/s13561-018-0195-5},
 keywords = {econ;phd},
 volume = {8},
 number = {1},
 issn = {2191-1991},
 journal = {Health Economics Review},
 doi = {10.1186/s13561-018-0195-5},
 howpublished = {refereed}
}

@article{Sciaini2018NLMR,
 author = {Sciaini, Marco and Fritsch, Matthias and Scherer, C{\'e}dric and Simpkins, Craig Eric},
 year = {2018},
 title = {{NLMR} and landscapetools : An integrated environment for simulating and modifying neutral landscape models in {R}},
 pages = {2240--2248},
 volume = {9},
 number = {11},
 issn = {2041210X},
 journal = {Methods in Ecology and Evolution},
 doi = {10.1111/2041-210X.13076},
 howpublished = {refereed}
}

@article{Simpkins2018Assessing,
 author = {Simpkins, Craig E. and Dennis, Todd E. and Etherington, Thomas R. and Perry, George L.W.},
 year = {2018},
 title = {Assessing the performance of common landscape connectivity metrics using a virtual ecologist approach},
 keywords = {ecol;postdoc},
 pages = {13--23},
 volume = {367},
 issn = {03043800},
 journal = {Ecological Modelling},
 doi = {10.1016/j.ecolmodel.2017.11.001},
 howpublished = {refereed}
}

@article{Simpkins2018Population,
 author = {Simpkins, Craig E. and Lee, Finnbar and Powers, Breanna and Anderson, Sandra and Asena, Quinn and Brock, James and Perry, George},
 year = {2018},
 title = {Population viability analyses in {New Zealand}: a review},
 keywords = {ecol;postdoc},
 pages = {107--115},
 volume = {42},
 number = {2},
 journal = {New Zealand Journal of Ecology},
 doi = {10.20417/nzjecol.42.32},
 howpublished = {refereed}
}

@article{Skevas2018Heterogeneity,
 author = {Skevas, Ioannis and Emvalomatis, Grigorios and Br{\"u}mmer, Bernhard},
 year = {2018},
 title = {Heterogeneity of Long-run Technical Efficiency of German Dairy Farms: A {Bayesian} Approach},
 keywords = {econ;phd},
 pages = {58--75},
 volume = {69},
 number = {1},
 issn = {0021857X},
 journal = {Journal of Agricultural Economics},
 doi = {10.1111/1477-9552.12231},
 howpublished = {refereed}
}

@article{Skevas2018Productivity,
 author = {Skevas, Ioannis and Emvalomatis, Grigorios and Br{\"u}mmer, Bernhard},
 year = {2018},
 title = {Productivity growth measurement and decomposition under a dynamic inefficiency specification: The case of {German} dairy farms},
 keywords = {econ;phd},
 pages = {250--261},
 volume = {271},
 number = {1},
 issn = {03772217},
 journal = {European Journal of Operational Research},
 doi = {10.1016/j.ejor.2018.04.050},
 howpublished = {refereed}
}

@article{Skevas2018TheEffect,
 author = {Skevas, Ioannis and Emvalomatis, Grigorios and Br{\"u}mmer, Bernhard},
 year = {2018},
 title = {The effect of farm characteristics on the persistence of technical inefficiency: a case study in {German} dairy farming},
 keywords = {econ;phd},
 pages = {3--25},
 volume = {45},
 number = {1},
 issn = {0165-1587},
 journal = {European Review of Agricultural Economics},
 doi = {10.1093/erae/jbx019},
 howpublished = {refereed}
}

@article{Thaden2018Structural,
 abstract = {In regression analyses of spatially structured data, it is common practice to introduce spatially correlated random effects into the regression model in order to reduce or even avoid unobserved variable bias in the estimation of other covariate effects. If besides the response the covariates are also spatially correlated, the spatial effects may confound the effect of the covariates or vice versa. In this case, the model fails to identify the true covariate effect due to multicollinearity. For highly collinear continuous covariates, path analysis and structural equation modeling techniques prove to be helpful to disentangle direct covariate effects from indirect covariate effects arising from correlation with other variables. This work discusses the applicability of these techniques in regression setups where spatial and covariate effects coincide at least partly and classical geoadditive models fail to separate these effects.},
 author = {Thaden, Hauke and Kneib, Thomas},
 year = {2018},
 title = {Structural Equation Models for Dealing With Spatial Confounding},
 url = {http://dx.doi.org/10.1080/00031305.2017.1305290},
 keywords = {phd;stat},
 pages = {239--252},
 volume = {72},
 number = {3},
 issn = {0003-1305},
 journal = {The American Statistician},
 doi = {10.1080/00031305.2017.1305290},
 howpublished = {refereed}
}

In regression analyses of spatially structured data, it is common practice to introduce spatially correlated random effects into the regression model in order to reduce or even avoid unobserved variable bias in the estimation of other covariate effects. If besides the response the covariates are also spatially correlated, the spatial effects may confound the effect of the covariates or vice versa. In this case, the model fails to identify the true covariate effect due to multicollinearity. For highly collinear continuous covariates, path analysis and structural equation modeling techniques prove to be helpful to disentangle direct covariate effects from indirect covariate effects arising from correlation with other variables. This work discusses the applicability of these techniques in regression setups where spatial and covariate effects coincide at least partly and classical geoadditive models fail to separate these effects.

@article{Yasmeen2018Relating,
 abstract = {In GAW20, we investigated the association of specific genetic regions of interest (ROIs) with log-transformed triglyceride (TG) levels following lipid-lowering medication using epigenetic and genetic markers. The goal was to incorporate kernels for cytosine-phosphate-guanine (CpG) markers and compare the kernels to a purely parametric model. Post-treatment TG levels were investigated for post-methylation data at CpG sites and region-specific SNPs and adjusted for pre-treatment TG levels and age, in independent individuals only (real data: n = 150; simulated data, replicate 84: n = 111). In both data sets, our single-CpG-marker results using kernels and linear regression were in good agreement. In the real data, we investigated the introns of the CPT1A gene previously reported as associated with TG levels as separate ROIs, and were able to find hints of an association of cg17058475 and cg00574958 with post-treatment TG levels. In the simulated data, we investigated a total of 10 regions, in which the 5 causal and 5 non-causal markers lie, respectively, with increased methylation variances, yielding plausible results for the 3 window sizes. Overall, this indicates that kernels for CpG markers are feasible. An interaction regression model for the causal SNP with the nearest CpG marker identified an effect for the SNPs with the three greatest heritabilities simulated. The simulation model assumed full SNP effect only for unmethylated regions decreasing to zero in the case of full methylation. Thus, in the context of a clear candidate setting, interaction between epigenetic and genetic data may enhance information, albeit nominally, even with small sample sizes. Relieving the burden of multiple testing, developing kernels further to analyze data from multiple omics jointly is well warranted.},
 author = {Yasmeen, Summaira and Burger, Patricia and Friedrichs, Stefanie and Papiol, Sergi and Bickeb{\"o}ller, Heike},
 year = {2018},
 title = {Relating drug response to epigenetic and genetic markers using a region-based kernel score test},
 keywords = {gen;phd},
 pages = {47},
 volume = {12},
 number = {Suppl 9},
 issn = {1753-6561},
 journal = {BMC Proceedings},
 doi = {10.1186/s12919-018-0154-5},
 howpublished = {refereed}
}

In GAW20, we investigated the association of specific genetic regions of interest (ROIs) with log-transformed triglyceride (TG) levels following lipid-lowering medication using epigenetic and genetic markers. The goal was to incorporate kernels for cytosine-phosphate-guanine (CpG) markers and compare the kernels to a purely parametric model. Post-treatment TG levels were investigated for post-methylation data at CpG sites and region-specific SNPs and adjusted for pre-treatment TG levels and age, in independent individuals only (real data: n = 150; simulated data, replicate 84: n = 111). In both data sets, our single-CpG-marker results using kernels and linear regression were in good agreement. In the real data, we investigated the introns of the CPT1A gene previously reported as associated with TG levels as separate ROIs, and were able to find hints of an association of cg17058475 and cg00574958 with post-treatment TG levels. In the simulated data, we investigated a total of 10 regions, in which the 5 causal and 5 non-causal markers lie, respectively, with increased methylation variances, yielding plausible results for the 3 window sizes. Overall, this indicates that kernels for CpG markers are feasible. An interaction regression model for the causal SNP with the nearest CpG marker identified an effect for the SNPs with the three greatest heritabilities simulated. The simulation model assumed full SNP effect only for unmethylated regions decreasing to zero in the case of full methylation. Thus, in the context of a clear candidate setting, interaction between epigenetic and genetic data may enhance information, albeit nominally, even with small sample sizes. Relieving the burden of multiple testing, developing kernels further to analyze data from multiple omics jointly is well warranted.

@article{Brink2017Genealogies,
 author = {Brink-Spalink, Rebekka and Smadi, Charline},
 year = {2017},
 title = {Genealogies of two linked neutral loci after a selective sweep in a large population of stochastically varying size},
 url = {http://dx.doi.org/10.1017/apr.2016.88},
 keywords = {gen;phd;stat},
 pages = {279--326},
 volume = {49},
 number = {01},
 issn = {0001-8678},
 journal = {Advances in Applied Probability},
 doi = {10.1017/apr.2016.88},
 howpublished = {refereed}
}

@article{Dislich2016AReview,
 abstract = {Oil palm plantations have expanded rapidly in recent decades. This large-scale land-use change has had great ecological, economic, and social impacts on both the areas converted to oil palm and their surroundings. However, research on the impacts of oil palm cultivation is scattered and patchy, and no clear overview exists. We address this gap through a systematic and comprehensive literature review of all ecosystem functions in oil palm plantations, including several (genetic, medicinal and ornamental resources, information functions) not included in previous systematic reviews. We compare ecosystem functions in oil palm plantations to those in forests, as the conversion of forest to oil palm is prevalent in the tropics. We find that oil palm plantations generally have reduced ecosystem functioning compared to forests: 11 out of 14 ecosystem functions show a net decrease in level of function. Some functions show decreases with potentially irreversible global impacts (e.g. reductions in gas and climate regulation, habitat and nursery functions, genetic resources, medicinal resources, and information functions). The most serious impacts occur when forest is cleared to establish new plantations, and immediately afterwards, especially on peat soils. To variable degrees, specific plantation management measures can prevent or reduce losses of some ecosystem functions (e.g. avoid illegal land clearing via fire, avoid draining of peat, use of integrated pest management, use of cover crops, mulch, and compost) and we highlight synergistic mitigation measures that can improve multiple ecosystem functions simultaneously. The~only ecosystem function which increases in oil palm plantations is, unsurprisingly, the production of marketable goods. Our review highlights numerous research gaps. In particular, there are significant gaps with respect to socio-cultural information functions. Further, there is a need for more empirical data on the importance of spatial and temporal scales, such as differences among plantations in different environments, of different sizes, and of different ages, as our review has identified examples where ecosystem functions vary spatially and temporally. Finally, more research is needed on developing management practices that can offset the losses of ecosystem functions. Our findings should stimulate research to address the identified gaps, and provide a foundation for more systematic research and discussion on ways to minimize the negative impacts and maximize the positive impacts of oil palm cultivation.},
 author = {Dislich, Claudia and Keyel, Alexander C. and {[...]} and Meyer, Katrin M. and Hess, Bastian and Kreft, Holger and Tscharntke, Teja and Wiegand, Kerstin},
 year = {2017},
 title = {A review of the ecosystem functions in oil palm plantations, using forests as a reference system},
 url = {http://dx.doi.org/10.1111/brv.12295},
 keywords = {ecol;postdoc},
 pages = {1539--1569},
 volume = {92},
 number = {3},
 journal = {Biological reviews of the Cambridge Philosophical Society},
 doi = {10.1111/brv.12295},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/27511961},
 howpublished = {refereed}
}

Oil palm plantations have expanded rapidly in recent decades. This large-scale land-use change has had great ecological, economic, and social impacts on both the areas converted to oil palm and their surroundings. However, research on the impacts of oil palm cultivation is scattered and patchy, and no clear overview exists. We address this gap through a systematic and comprehensive literature review of all ecosystem functions in oil palm plantations, including several (genetic, medicinal and ornamental resources, information functions) not included in previous systematic reviews. We compare ecosystem functions in oil palm plantations to those in forests, as the conversion of forest to oil palm is prevalent in the tropics. We find that oil palm plantations generally have reduced ecosystem functioning compared to forests: 11 out of 14 ecosystem functions show a net decrease in level of function. Some functions show decreases with potentially irreversible global impacts (e.g. reductions in gas and climate regulation, habitat and nursery functions, genetic resources, medicinal resources, and information functions). The most serious impacts occur when forest is cleared to establish new plantations, and immediately afterwards, especially on peat soils. To variable degrees, specific plantation management measures can prevent or reduce losses of some ecosystem functions (e.g. avoid illegal land clearing via fire, avoid draining of peat, use of integrated pest management, use of cover crops, mulch, and compost) and we highlight synergistic mitigation measures that can improve multiple ecosystem functions simultaneously. The only ecosystem function which increases in oil palm plantations is, unsurprisingly, the production of marketable goods. Our review highlights numerous research gaps. In particular, there are significant gaps with respect to socio-cultural information functions. Further, there is a need for more empirical data on the importance of spatial and temporal scales, such as differences among plantations in different environments, of different sizes, and of different ages, as our review has identified examples where ecosystem functions vary spatially and temporally. Finally, more research is needed on developing management practices that can offset the losses of ecosystem functions. Our findings should stimulate research to address the identified gaps, and provide a foundation for more systematic research and discussion on ways to minimize the negative impacts and maximize the positive impacts of oil palm cultivation.

@article{friedrichs2017pathway,
 author = {Friedrichs, Stefanie and Manitz, Juliane and {[...]} and Kneib, Thomas and Bickeb{\"o}ller, Heike and Hofner, Benjamin},
 year = {2017},
 title = {Pathway-Based Kernel Boosting for the Analysis of Genome-Wide Association Studies},
 url = {http://dx.doi.org/10.1155/2017/6742763},
 keywords = {gen;phd;postdoc},
 pages = {1--17},
 volume = {2017},
 issn = {1748-670X},
 journal = {Computational and Mathematical Methods in Medicine},
 doi = {10.1155/2017/6742763},
 howpublished = {refereed}
}

@article{Gao2017Incorporating,
 abstract = {Today, genomic prediction (GP) is an established technology in plant and animal breeding programs. Current standard methods are purely based on statistical considerations but do not make use of the abundant biological knowledge, which is easily available from public databases. Major questions that have to be answered before biological prior information can be used routinely in GP approaches are which types of information can be used, and at which points they can be incorporated into prediction methods. In this study, we propose a novel strategy to incorporate gene annotation into GP of complex phenotypes by defining haploblocks according to gene positions. Haplotype effects are then modeled as categorical or as numerical allele dosage variables. The underlying concept of this approach is to build the statistical model on variables representing the biologically functional units. We evaluate the new methods with data from a heterogeneous stock mouse population, the Drosophila Genetic Reference Panel (DGRP), and a rice breeding population from the Rice Diversity Panel. Our results show that using gene annotation to define haploblocks often leads to a comparable, but for some traits to a higher, predictive ability compared to SNP-based models or to haplotype models that do not use gene annotation information. Modeling gene interaction effects can further improve predictive ability. We also illustrate that the additional use of markers that have not been mapped to any gene in a second separate relatedness matrix does in many cases not lead to a relevant additional increase in predictive ability when the first matrix is based on haploblocks defined with gene annotation data, suggesting that intergenic markers only provide redundant information on the considered data sets. Therefore, gene annotation information seems to be appropriate to perceive the importance of DNA segments. Finally, we discuss the effects of gene annotation quality, marker density, and linkage disequilibrium on the performance of the new methods. To our knowledge, this is the first work that incorporates epistatic interaction or gene annotation into haplotype-based prediction approaches.},
 author = {Gao, Ning and Martini, Johannes W. R. and Zhang, Zhe and Yuan, Xiaolong and Zhang, Hao and Simianer, Henner and Li, Jiaqi},
 year = {2017},
 title = {Incorporating Gene Annotation into Genomic Prediction of Complex Phenotypes},
 keywords = {gen;phd},
 pages = {489--501},
 volume = {207},
 number = {2},
 issn = {1943-2631},
 journal = {Genetics},
 doi = {10.1534/genetics.117.300198},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/28839043},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5629318},
 howpublished = {refereed}
}

Today, genomic prediction (GP) is an established technology in plant and animal breeding programs. Current standard methods are purely based on statistical considerations but do not make use of the abundant biological knowledge, which is easily available from public databases. Major questions that have to be answered before biological prior information can be used routinely in GP approaches are which types of information can be used, and at which points they can be incorporated into prediction methods. In this study, we propose a novel strategy to incorporate gene annotation into GP of complex phenotypes by defining haploblocks according to gene positions. Haplotype effects are then modeled as categorical or as numerical allele dosage variables. The underlying concept of this approach is to build the statistical model on variables representing the biologically functional units. We evaluate the new methods with data from a heterogeneous stock mouse population, the Drosophila Genetic Reference Panel (DGRP), and a rice breeding population from the Rice Diversity Panel. Our results show that using gene annotation to define haploblocks often leads to a comparable, but for some traits to a higher, predictive ability compared to SNP-based models or to haplotype models that do not use gene annotation information. Modeling gene interaction effects can further improve predictive ability. We also illustrate that the additional use of markers that have not been mapped to any gene in a second separate relatedness matrix does in many cases not lead to a relevant additional increase in predictive ability when the first matrix is based on haploblocks defined with gene annotation data, suggesting that intergenic markers only provide redundant information on the considered data sets. Therefore, gene annotation information seems to be appropriate to perceive the importance of DNA segments. Finally, we discuss the effects of gene annotation quality, marker density, and linkage disequilibrium on the performance of the new methods. To our knowledge, this is the first work that incorporates epistatic interaction or gene annotation into haplotype-based prediction approaches.

@article{Ha2017AReaction,
 abstract = {Due to the discrepancy of the high energy demand for rapidly increasing milk production and limited feed intake in the transition period around parturition, dairy cows require considerable metabolic adaptations. We hypothesize that some cows are genetically less suited to cope with these metabolic needs than others, leading to adverse follow-up effects on longevity. To test this, we designed a reaction norm model in which functional lifetime was linked to the metabolic challenge in the beginning of the first lactation. As challenge variables, we used either the sum of milk yield or the accumulated fat-to-protein ratio of the first 3 test-days ({\textless}120 d in milk), pre-adjusted for herd-test-day variance. We defined a random regression sire model, in which a random slope was estimated for each sire to assess whether a bull had robust (neutral or positive slopes) or non-robust (negative slopes) daughters. We fitted the model to data of $\sim$580,000 daughters of $\sim$5,000 Brown Swiss bulls with suitable observations available ($\geq$10 daughters per bull). To validate our proposed model and assess the reliability of the estimated (co)variance components, we conducted an extensive bootstrap approach. For both challenge variables, we found the sire variance for the slope of the random regression to be significantly different from zero, suggesting a genetic component for metabolic adaptability. The results of the study show that the ability to cope with metabolic stress in the transition period has a genetic component, which can be used to breed metabolically robust dairy cows.},
 author = {Ha, Ngoc-Thuy and Sharifi, A. R. and Heise, J. and Schlather, M. and Schnyder, U. and Gross, J. J. and Schmitz-Hsu, F. and Bruckmaier, R. M. and Simianer, H.},
 year = {2017},
 title = {A reaction norm sire model to study the effect of metabolic challenge in early lactation on the functional longevity of dairy cows},
 keywords = {gen;postdoc},
 pages = {3742--3753},
 volume = {100},
 number = {5},
 journal = {Journal of Dairy Science},
 doi = {10.3168/jds.2016-12031},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/28284692},
 howpublished = {refereed}
}

Due to the discrepancy of the high energy demand for rapidly increasing milk production and limited feed intake in the transition period around parturition, dairy cows require considerable metabolic adaptations. We hypothesize that some cows are genetically less suited to cope with these metabolic needs than others, leading to adverse follow-up effects on longevity. To test this, we designed a reaction norm model in which functional lifetime was linked to the metabolic challenge in the beginning of the first lactation. As challenge variables, we used either the sum of milk yield or the accumulated fat-to-protein ratio of the first 3 test-days (\textless120 d in milk), pre-adjusted for herd-test-day variance. We defined a random regression sire model, in which a random slope was estimated for each sire to assess whether a bull had robust (neutral or positive slopes) or non-robust (negative slopes) daughters. We fitted the model to data of $∼$580,000 daughters of $∼$5,000 Brown Swiss bulls with suitable observations available ($≥$10 daughters per bull). To validate our proposed model and assess the reliability of the estimated (co)variance components, we conducted an extensive bootstrap approach. For both challenge variables, we found the sire variance for the slope of the random regression to be significantly different from zero, suggesting a genetic component for metabolic adaptability. The results of the study show that the ability to cope with metabolic stress in the transition period has a genetic component, which can be used to breed metabolically robust dairy cows.

@article{Ha2017Genetische,
 author = {Ha, Ngoc-Thuy and Gross, J.J. and Sharifi, A.R. and Schlather, M. and Dr{\"o}gem{\"u}ller, C. and Schnyder, U. and Schmitz-Hsu, F. and Bruckmaier, R.M. and Simianer, H.},
 year = {2017},
 title = {Genetische {A}nalyse der metabolischen {A}daptation von {M}ilchk{\"u}hen in der {F}r{\"u}hlaktation},
 pages = {48--60},
 journal = {Z{\"u}chtungskunde},
 howpublished = {refereed}
}

@article{Ha2017Liver,
 abstract = {During early lactation, dairy cows experience a severe metabolic load often resulting in the development of various diseases. The inevitable deficiency in nutrients and energy at the onset of lactation requires an optimal adaptation of the hepatic metabolism to overcome metabolic stress. We conducted a whole-liver transcriptome analysis for the transition cow to identify novel factors crucial for metabolic adaptation. Liver samples were obtained from 6 Red Holstein dairy cows (parity 2 to 7, mean $\pm$ standard deviation: 3.7 $\pm$ 2.3) at 3 time points: T1 = 22 $\pm$ 4 d antepartum, T2 = 10 $\pm$ 2 d postpartum, and T3 = 17 $\pm$ 2 d postpartum. Using RNA sequencing (RNA-seq), we studied the transcriptomic profile of the transition cow before and after parturition. We performed a differential gene expression analysis (DGEA) and gene-set enrichment analysis (GSEA) for biological processes (gene ontology, GO) and pathways (Kyoto Encyclopedia of Genes and Genomes, KEGG). Among the 10,186 expressed genes, we discovered 1,063 differentially expressed genes (false discovery rate = 5{\%}). The GSEA revealed 16 biological processes and 7 pathways significantly (false discovery rate = 5{\%}) associated with the hepatic changes of the transition cow. Our results confirm that major hepatic changes are related to energy mobilization after parturition; in particular, they are related to fatty acid oxidation/metabolism, cholesterol metabolism, and gluconeogenesis. Using the STRING database (https://string-db.org/), we investigated interactions between significant genes and identified 9 key genes (CYP7A1, APOA1, CREM, LOC522146, CYP2C87, HMGCR, FDFT1, SGLE, and CYP26A1) through which the different processes involved in the metabolic adaptation interact. Comparing our main results with the literature, we could identify further genes that have not yet been associated with the transition period (e.g., CPT1B, ADIPOR2, LEPR, CREB3L3, and CCND1) and that are mainly involved in processes controlled by AMP-activated protein kinase, an important regulator of energy homeostasis.},
 author = {Ha, Ngoc-Thuy and Dr{\"o}gem{\"u}ller, C. and Reimer, C. and Schmitz-Hsu, F. and Bruckmaier, R. M. and Simianer, H. and Gross, J. J.},
 year = {2017},
 title = {Liver transcriptome analysis reveals important factors involved in the metabolic adaptation of the transition cow},
 keywords = {gen;postdoc},
 pages = {9311--9323},
 volume = {100},
 number = {11},
 journal = {Journal of Dairy Science},
 doi = {10.3168/jds.2016-12454},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/28865861},
 howpublished = {refereed}
}

During early lactation, dairy cows experience a severe metabolic load often resulting in the development of various diseases. The inevitable deficiency in nutrients and energy at the onset of lactation requires an optimal adaptation of the hepatic metabolism to overcome metabolic stress. We conducted a whole-liver transcriptome analysis for the transition cow to identify novel factors crucial for metabolic adaptation. Liver samples were obtained from 6 Red Holstein dairy cows (parity 2 to 7, mean $±$ standard deviation: 3.7 $±$ 2.3) at 3 time points: T1 = 22 $±$ 4 d antepartum, T2 = 10 $±$ 2 d postpartum, and T3 = 17 $±$ 2 d postpartum. Using RNA sequencing (RNA-seq), we studied the transcriptomic profile of the transition cow before and after parturition. We performed a differential gene expression analysis (DGEA) and gene-set enrichment analysis (GSEA) for biological processes (gene ontology, GO) and pathways (Kyoto Encyclopedia of Genes and Genomes, KEGG). Among the 10,186 expressed genes, we discovered 1,063 differentially expressed genes (false discovery rate = 5%). The GSEA revealed 16 biological processes and 7 pathways significantly (false discovery rate = 5%) associated with the hepatic changes of the transition cow. Our results confirm that major hepatic changes are related to energy mobilization after parturition; in particular, they are related to fatty acid oxidation/metabolism, cholesterol metabolism, and gluconeogenesis. Using the STRING database (https://string-db.org/), we investigated interactions between significant genes and identified 9 key genes (CYP7A1, APOA1, CREM, LOC522146, CYP2C87, HMGCR, FDFT1, SGLE, and CYP26A1) through which the different processes involved in the metabolic adaptation interact. Comparing our main results with the literature, we could identify further genes that have not yet been associated with the transition period (e.g., CPT1B, ADIPOR2, LEPR, CREB3L3, and CCND1) and that are mainly involved in processes controlled by AMP-activated protein kinase, an important regulator of energy homeostasis.

@article{Hambuckers2017HowCan,
 author = {Hambuckers, Julien and Dauvrin, A. and Trolliet, F. and Evrard, Q. and Forget, P. M. and Hambuckers, A.},
 year = {2017},
 title = {How can seed removal rates of zoochoric tree species be assessed quickly and accurately?},
 url = {http://dx.doi.org/10.1016/j.foreco.2017.07.042},
 keywords = {ecol;postdoc;stat},
 pages = {152--160},
 volume = {403},
 issn = {03781127},
 journal = {Forest Ecology and Management},
 doi = {10.1016/j.foreco.2017.07.042},
 howpublished = {refereed}
}

@article{LibranEmbid2017Effects,
 author = {Libr{\'a}n-Embid, Felipe and Coster, Greet De and Metzger, Jean Paul},
 year = {2017},
 title = {Effects of bird and bat exclusion on coffee pest control at multiple spatial scales},
 url = {https://link.springer.com/content/pdf/10.1007/s10980-017-0555-2.pdf},
 keywords = {ecol;phd},
 pages = {1907--1920},
 volume = {32},
 number = {9},
 issn = {1572-9761},
 journal = {Landscape Ecology},
 doi = {10.1007/s10980-017-0555-2},
 file = {2edb4d8d-f6e0-4834-b3bc-a67b27a27509:C\:\\Users\\Barbara\\AppData\\Local\\Swiss Academic Software\\Citavi 6\\ProjectCache\\f4q6lsjyptp8cy52jyikg6ejy8o6t76o6gw6o347n6vucmn\\Citavi Attachments\\2edb4d8d-f6e0-4834-b3bc-a67b27a27509.pdf:pdf},
 howpublished = {refereed}
}

@article{Manitz2017Source,
 author = {Manitz, Juliane and Harbering, Jonas and Schmidt, Marie and Kneib, Thomas and Sch{\"o}bel, Anita},
 year = {2017},
 title = {Source estimation for propagation processes on complex networks with an application to delays in public transportation systems},
 url = {http://dx.doi.org/10.1111/rssc.12176},
 keywords = {postdoc;stat},
 pages = {521--536},
 volume = {66},
 number = {3},
 issn = {00359254},
 journal = {Journal of the Royal Statistical Society: Series C (Applied Statistics)},
 doi = {10.1111/rssc.12176},
 howpublished = {refereed}
}

@article{Martini2017Genomic,
 abstract = {BACKGROUND

Epistasis marker effect models incorporating products of marker values as predictor variables in a linear regression approach (extended GBLUP, EGBLUP) have been assessed as potentially beneficial for genomic prediction, but their performance depends on marker coding. Although this fact has been recognized in literature, the nature of the problem has not been thoroughly investigated so far.

RESULTS

We illustrate how the choice of marker coding implicitly specifies the model of how effects of certain allele combinations at different loci contribute to the phenotype, and investigate coding-dependent properties of EGBLUP. Moreover, we discuss an alternative categorical epistasis model (CE) eliminating undesired properties of EGBLUP and show that the CE model can improve predictive ability. Finally, we demonstrate that the coding-dependent performance of EGBLUP offers the possibility to incorporate prior experimental information into the prediction method by adapting the coding to already available phenotypic records on other traits.

CONCLUSION

Based on our results, for EGBLUP, a symmetric coding {-1,1} or {-1,0,1} should be preferred, whereas a standardization using allele frequencies should be avoided. Moreover, CE can be a valuable alternative since it does not possess the undesired theoretical properties of EGBLUP. However, which model performs best will depend on characteristics of the data and available prior information. Data from previous experiments can for instance be incorporated into the marker coding of EGBLUP.},
 author = {Martini, Johannes W. R. and Gao, Ning and Cardoso, Diercles F. and Wimmer, Valentin and Erbe, Malena and Cantet, Rodolfo J. C. and Simianer, Henner},
 year = {2017},
 title = {Genomic prediction with epistasis models: on the marker-coding-dependent performance of the extended {GBLUP} and properties of the categorical epistasis model {(CE)}},
 keywords = {gen;phd},
 pages = {3},
 volume = {18},
 number = {1},
 journal = {BMC Bioinformatics},
 doi = {10.1186/s12859-016-1439-1},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/28049412},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5209948},
 howpublished = {refereed}
}

BACKGROUND Epistasis marker effect models incorporating products of marker values as predictor variables in a linear regression approach (extended GBLUP, EGBLUP) have been assessed as potentially beneficial for genomic prediction, but their performance depends on marker coding. Although this fact has been recognized in literature, the nature of the problem has not been thoroughly investigated so far. RESULTS We illustrate how the choice of marker coding implicitly specifies the model of how effects of certain allele combinations at different loci contribute to the phenotype, and investigate coding-dependent properties of EGBLUP. Moreover, we discuss an alternative categorical epistasis model (CE) eliminating undesired properties of EGBLUP and show that the CE model can improve predictive ability. Finally, we demonstrate that the coding-dependent performance of EGBLUP offers the possibility to incorporate prior experimental information into the prediction method by adapting the coding to already available phenotypic records on other traits. CONCLUSION Based on our results, for EGBLUP, a symmetric coding -1,1 or -1,0,1 should be preferred, whereas a standardization using allele frequencies should be avoided. Moreover, CE can be a valuable alternative since it does not possess the undesired theoretical properties of EGBLUP. However, which model performs best will depend on characteristics of the data and available prior information. Data from previous experiments can for instance be incorporated into the marker coding of EGBLUP.

@article{Rosenberger2017Gene,
 abstract = {Gene-set analysis (GSA) is an approach using the results of single-marker genome-wide association studies when investigating pathways as a whole with respect to the genetic basis of a disease. We performed a meta-analysis of seven GSAs for lung cancer, applying the method META-GSA. Overall, the information taken from 11,365 cases and 22,505 controls from within the TRICL/ILCCO consortia was used to investigate a total of 234 pathways from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. META-GSA reveals the systemic lupus erythematosus KEGG pathway hsa05322, driven by the gene region 6p21-22, as also implicated in lung cancer (p = 0.0306). This gene region is known to be associated with squamous cell lung carcinoma. The most important genes driving the significance of this pathway belong to the genomic areas HIST1-H4L, -1BN, -2BN, -H2AK, -H4K and C2/C4A/C4B. Within these areas, the markers most significantly associated with LC are rs13194781 (located within HIST12BN) and rs1270942 (located between C2 and C4A). We have discovered a pathway currently marked as specific to systemic lupus erythematosus as being significantly implicated in lung cancer. The gene region 6p21-22 in this pathway appears to be more extensively associated with lung cancer than previously assumed. Given wide-stretched linkage disequilibrium to the area APOM/BAG6/MSH5, there is currently simply not enough information or evidence to conclude whether the potential pleiotropy of lung cancer and systemic lupus erythematosus is spurious, biological, or mediated. Further research into this pathway and gene region will be necessary.},
 author = {Rosenberger, Albert and Sohns, Melanie and Friedrichs, Stefanie and {[...]} and Bickeb{\"o}ller, Heike},
 year = {2017},
 title = {Gene-set meta-analysis of lung cancer identifies pathway related to systemic lupus erythematosus},
 url = {http://dx.doi.org/10.1371/journal.pone.0173339},
 keywords = {gen;phd},
 pages = {e0173339+},
 volume = {12},
 number = {3},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0173339},
 howpublished = {refereed}
}

Gene-set analysis (GSA) is an approach using the results of single-marker genome-wide association studies when investigating pathways as a whole with respect to the genetic basis of a disease. We performed a meta-analysis of seven GSAs for lung cancer, applying the method META-GSA. Overall, the information taken from 11,365 cases and 22,505 controls from within the TRICL/ILCCO consortia was used to investigate a total of 234 pathways from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. META-GSA reveals the systemic lupus erythematosus KEGG pathway hsa05322, driven by the gene region 6p21-22, as also implicated in lung cancer (p = 0.0306). This gene region is known to be associated with squamous cell lung carcinoma. The most important genes driving the significance of this pathway belong to the genomic areas HIST1-H4L, -1BN, -2BN, -H2AK, -H4K and C2/C4A/C4B. Within these areas, the markers most significantly associated with LC are rs13194781 (located within HIST12BN) and rs1270942 (located between C2 and C4A). We have discovered a pathway currently marked as specific to systemic lupus erythematosus as being significantly implicated in lung cancer. The gene region 6p21-22 in this pathway appears to be more extensively associated with lung cancer than previously assumed. Given wide-stretched linkage disequilibrium to the area APOM/BAG6/MSH5, there is currently simply not enough information or evidence to conclude whether the potential pleiotropy of lung cancer and systemic lupus erythematosus is spurious, biological, or mediated. Further research into this pathway and gene region will be necessary.

@article{Simpkins2017Understanding,
 author = {Simpkins, Craig E. and Perry, George L.W.},
 year = {2017},
 title = {Understanding the impacts of temporal variability on estimates of landscape connectivity},
 pages = {243--248},
 volume = {83},
 issn = {1470160X},
 journal = {Ecological Indicators},
 doi = {10.1016/j.ecolind.2017.08.008},
 howpublished = {refereed}
}

@article{Thaden2017Integrating,
 author = {Thaden, Hauke and Pata, Mar$\backslash$'{\i}a P. and Klein, Nadja and Cadarso-Su{\'a}rez, Carmen and Kneib, Thomas},
 year = {2017},
 title = {Integrating multivariate conditionally autoregressive spatial priors into recursive bivariate models for analyzing environmental sensitivity of mussels},
 url = {http://dx.doi.org/10.1016/j.spasta.2017.07.005},
 keywords = {phd;stat},
 pages = {419--433},
 volume = {22},
 issn = {22116753},
 journal = {Spatial Statistics},
 doi = {10.1016/j.spasta.2017.07.005},
 howpublished = {refereed}
}

@article{Tifaoui2017Temporary,
 abstract = {We hypothesize that temporary sales might bias the results of vertical price transmission analysis towards findings of asymmetric price transmission. We test this hypothesis using a scanner dataset of retail butter prices in Germany. To this end we first use filters to identify underlying reference retail prices and thus remove temporary sales prices. We then compare estimates of vertical price transmission from the wholesale to the retail level that are generated with raw and with filtered retail prices. Our results confirm that temporary sales prices increase the speed and asymmetry of vertical price transmission. These results add a potential cause of asymmetry to those, such as market power, that have already been identified in the literature. [EconLit citations: C22; L10; Q11].},
 author = {Tifaoui, Said and von Cramon-Taubadel, Stephan},
 year = {2017},
 title = {Temporary Sales Prices and Asymmetric Price Transmission},
 url = {http://dx.doi.org/10.1002/agr.21465},
 keywords = {econ;phd},
 pages = {85--97},
 volume = {33},
 number = {1},
 issn = {0742-4477},
 journal = {Agribusiness},
 doi = {10.1002/agr.21465},
 howpublished = {refereed}
}

We hypothesize that temporary sales might bias the results of vertical price transmission analysis towards findings of asymmetric price transmission. We test this hypothesis using a scanner dataset of retail butter prices in Germany. To this end we first use filters to identify underlying reference retail prices and thus remove temporary sales prices. We then compare estimates of vertical price transmission from the wholesale to the retail level that are generated with raw and with filtered retail prices. Our results confirm that temporary sales prices increase the speed and asymmetry of vertical price transmission. These results add a potential cause of asymmetry to those, such as market power, that have already been identified in the literature. [EconLit citations: C22; L10; Q11].

@article{Friedrichs2016Filtering,
 author = {Friedrichs, Stefanie and Malzahn, D{\"o}rthe and Pugh, Elizabeth W. and Almeida, Marcio and Liu, Xiao Q. and Bailey, Julia N.},
 year = {2016},
 title = {Filtering genetic variants and placing informative priors based on putative biological function},
 url = {http://dx.doi.org/10.1186/s12863-015-0313-x},
 keywords = {gen;phd},
 volume = {17},
 number = {S2},
 issn = {1471-2156},
 journal = {BMC Genetics},
 doi = {10.1186/s12863-015-0313-x},
 howpublished = {refereed}
}

@article{Hambuckers2016Estimating,
 abstract = {In this paper, we provide a novel way to estimate the out-of-sample predictive ability of a trading rule. Usually, this ability is estimated using a sample-splitting scheme, true out-of-sample data being rarely available. We argue that this method makes poor use of the available data and creates data-mining possibilities. Instead, we introduce an alternative.632 bootstrap approach. This method enables building in-sample and out-of-sample bootstrap datasets that do not overlap but exhibit the same time dependencies. We show in a simulation study that this technique drastically reduces the mean squared error of the estimated predictive ability. We illustrate our methodology on IBM, MSFT and DJIA stock prices, where we compare 11 trading rules specifications. For the considered datasets, two different filter rule specifications have the highest out-of-sample mean excess returns. However, all tested rules cannot beat a simple buy-and-hold strategy when trading at a daily frequency. Copyright {\copyright} 2015 John Wiley {\&} Sons, Ltd.},
 author = {Hambuckers, Julien and Heuchenne, C{\'e}dric},
 year = {2016},
 title = {Estimating the Out-of-Sample Predictive Ability of Trading Rules: A Robust Bootstrap Approach},
 url = {http://dx.doi.org/10.1002/for.2380},
 keywords = {postdoc;stat},
 pages = {347--372},
 volume = {35},
 number = {4},
 issn = {02776693},
 journal = {Journal of Forecasting},
 doi = {10.1002/for.2380},
 howpublished = {refereed}
}

In this paper, we provide a novel way to estimate the out-of-sample predictive ability of a trading rule. Usually, this ability is estimated using a sample-splitting scheme, true out-of-sample data being rarely available. We argue that this method makes poor use of the available data and creates data-mining possibilities. Instead, we introduce an alternative.632 bootstrap approach. This method enables building in-sample and out-of-sample bootstrap datasets that do not overlap but exhibit the same time dependencies. We show in a simulation study that this technique drastically reduces the mean squared error of the estimated predictive ability. We illustrate our methodology on IBM, MSFT and DJIA stock prices, where we compare 11 trading rules specifications. For the considered datasets, two different filter rule specifications have the highest out-of-sample mean excess returns. However, all tested rules cannot beat a simple buy-and-hold strategy when trading at a daily frequency. Copyright © 2015 John Wiley & Sons, Ltd.

@article{Hambuckers2017ARobust,
 author = {Hambuckers, Julien and Heuchenne, C.},
 year = {2016},
 title = {A robust statistical approach to select adequate error distributions for financial returns},
 url = {http://dx.doi.org/10.1080/02664763.2016.1165803},
 keywords = {postdoc;stat},
 pages = {137--161},
 volume = {44},
 number = {1},
 issn = {0266-4763},
 journal = {Journal of Applied Statistics},
 doi = {10.1080/02664763.2016.1165803},
 howpublished = {refereed}
}

@article{Herwartz2016Modelling,
 author = {Herwartz, Helmut and Klein, Nadja and Strumann, Christoph},
 year = {2016},
 title = {Modelling Hospital Admission and Length of Stay by Means of Generalised Count Data Models},
 url = {http://dx.doi.org/10.1002/jae.2454},
 keywords = {econ;phd;stat},
 pages = {1159--1182},
 volume = {31},
 number = {6},
 issn = {08837252},
 journal = {Journal of Applied Econometrics},
 doi = {10.1002/jae.2454},
 howpublished = {refereed}
}

@article{Keyel2016Spatial,
 author = {Keyel, Alexander C. and Gerstenlauer, Jakob L. K. and Wiegand, Kerstin},
 year = {2016},
 title = {{SpatialDemography}: a spatially explicit, stage-structured, metacommunity model},
 url = {http://dx.doi.org/10.1111/ecog.02295},
 keywords = {ecol;postdoc},
 pages = {1129--1137},
 volume = {39},
 number = {11},
 issn = {09067590},
 journal = {Ecography},
 doi = {10.1111/ecog.02295},
 howpublished = {refereed}
}

@article{Keyel2016Validating,
 author = {Keyel, Alexander C. and Wiegand, Kerstin},
 year = {2016},
 title = {Validating the use of unique trait combinations for measuring multivariate functional richness},
 url = {http://dx.doi.org/10.1111/2041-210x.12558},
 keywords = {ecol;postdoc},
 pages = {929--936},
 volume = {7},
 number = {8},
 issn = {2041210X},
 journal = {Methods in Ecology and Evolution},
 doi = {10.1111/2041-210x.12558},
 howpublished = {refereed}
}

@article{Klein2016Scale,
 author = {Klein, Nadja and Kneib, Thomas},
 year = {2016},
 title = {Scale-Dependent Priors for Variance Parameters in Structured Additive Distributional Regression},
 url = {http://dx.doi.org/10.1214/15-ba983},
 keywords = {phd;stat},
 pages = {1071--1106},
 volume = {11},
 number = {4},
 issn = {1936-0975},
 journal = {Bayesian Analysis},
 doi = {10.1214/15-ba983},
 howpublished = {refereed}
}

@article{Kormann2016Corridors,
 abstract = {Tropical biodiversity and associated ecosystem functions have become heavily eroded through habitat loss. Animal-mediated pollination is required in more than 94{\%} of higher tropical plant species and 75{\%} of the world's leading food crops, but it remains unclear if corridors avert deforestation-driven pollination breakdown in fragmented tropical landscapes. Here, we used manipulative resource experiments and field observations to show that corridors functionally connect neotropical forest fragments for forest-associated hummingbirds and increase pollen transfer. Further, corridors boosted forest-associated pollinator availability in fragments by 14.3 times compared with unconnected equivalents, increasing overall pollination success. Plants in patches without corridors showed pollination rates equal to bagged control flowers, indicating pollination failure in isolated fragments. This indicates, for the first time, that corridors benefit tropical forest ecosystems beyond boosting local species richness, by functionally connecting mutualistic network partners. We conclude that small-scale adjustments to landscape configuration safeguard native pollinators and associated pollination services in tropical forest landscapes.},
 author = {Kormann, Urs and Scherber, Christoph and Tscharntke, Teja and Klein, Nadja and Larbig, Manuel and Valente, Jonathon J. and Hadley, Adam S. and Betts, Matthew G.},
 year = {2016},
 title = {Corridors restore animal-mediated pollination in fragmented tropical forest landscapes},
 keywords = {ecol;phd},
 pages = {20152347},
 volume = {283},
 number = {1823},
 issn = {1471-2954},
 journal = {Proceedings of the Royal Society B},
 doi = {10.1098/rspb.2015.2347},
 file = {http://dx.doi.org/10.1098/rspb.2015.2347},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/26817765},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4795022},
 howpublished = {refereed}
}

Tropical biodiversity and associated ecosystem functions have become heavily eroded through habitat loss. Animal-mediated pollination is required in more than 94% of higher tropical plant species and 75% of the world's leading food crops, but it remains unclear if corridors avert deforestation-driven pollination breakdown in fragmented tropical landscapes. Here, we used manipulative resource experiments and field observations to show that corridors functionally connect neotropical forest fragments for forest-associated hummingbirds and increase pollen transfer. Further, corridors boosted forest-associated pollinator availability in fragments by 14.3 times compared with unconnected equivalents, increasing overall pollination success. Plants in patches without corridors showed pollination rates equal to bagged control flowers, indicating pollination failure in isolated fragments. This indicates, for the first time, that corridors benefit tropical forest ecosystems beyond boosting local species richness, by functionally connecting mutualistic network partners. We conclude that small-scale adjustments to landscape configuration safeguard native pollinators and associated pollination services in tropical forest landscapes.

@article{Madeira2016Spillover,
 author = {Madeira, Filipe and Tscharntke, Teja and Elek, Zolt{\'a}n and Kormann, Urs and Pons, Xavier and R{\"o}sch, Verena and Samu, Ferenc and Scherber, Christoph and Bat{\'a}ry, P{\'e}ter},
 year = {2016},
 title = {Spillover of arthropods from cropland to protected calcareous grassland --- the neighbouring habitat matters},
 url = {http://dx.doi.org/10.1016/j.agee.2016.10.012},
 keywords = {ecol;phd},
 pages = {127--133},
 volume = {235},
 issn = {01678809},
 journal = {Agriculture, Ecosystems {\&} Environment},
 doi = {10.1016/j.agee.2016.10.012},
 howpublished = {refereed}
}

@article{Maerz2016Analysing,
 author = {M{\"a}rz, Alexander and Klein, Nadja and Kneib, Thomas and Musshoff, Oliver},
 year = {2016},
 title = {Analysing farmland rental rates using {Bayesian} geoadditive quantile regression},
 url = {http://dx.doi.org/10.1093/erae/jbv028},
 keywords = {phd;stat},
 pages = {663--698},
 volume = {43},
 number = {4},
 issn = {0165-1587},
 journal = {European Review of Agricultural Economics},
 doi = {10.1093/erae/jbv028},
 howpublished = {refereed}
}

@article{Malzahn2016Comparing,
 author = {Malzahn, D{\"o}rthe and Friedrichs, Stefanie and Bickeb{\"o}ller, Heike},
 year = {2016},
 title = {Comparing strategies for combined testing of rare and common variants in whole sequence and genome-wide genotype data},
 url = {http://dx.doi.org/10.1186/s12919-016-0042-9},
 keywords = {gen;phd},
 volume = {10},
 number = {S7},
 issn = {1753-6561},
 journal = {BMC Proceedings},
 doi = {10.1186/s12919-016-0042-9},
 howpublished = {refereed}
}

@article{Martini2016Epistasis,
 abstract = {KEY MESSAGE

Models based on additive marker effects and on epistatic interactions can be translated into genomic relationship models. This equivalence allows to perform predictions based on complex gene interaction models and reduces computational effort significantly. In the theory of genome-assisted prediction, the equivalence of a linear model based on independent and identically normally distributed marker effects and a model based on multivariate Gaussian distributed breeding values with genomic relationship as covariance matrix is well known. In this work, we demonstrate equivalences of marker effect models incorporating epistatic interactions and corresponding mixed models based on relationship matrices and show how to exploit these equivalences computationally for genome-assisted prediction. In particular, we show how models with epistatic interactions of higher order (e.g., three-factor interactions) translate into linear models with certain covariance matrices and demonstrate how to construct epistatic relationship matrices for the linear mixed model, if we restrict the model to interactions defined a priori. We illustrate the practical relevance of our results with a publicly available data set on grain yield of wheat lines growing in four different environments. For this purpose, we select important interactions in one environment and use this knowledge on the network of interactions to increase predictive ability of grain yield under other environmental conditions. Our results provide a guide for building relationship matrices based on knowledge on the structure of trait-related gene networks.},
 author = {Martini, Johannes W. R. and Wimmer, Valentin and Erbe, Malena and Simianer, Henner},
 year = {2016},
 title = {Epistasis and covariance: how gene interaction translates into genomic relationship},
 keywords = {gen;phd},
 pages = {963--976},
 volume = {129},
 number = {5},
 journal = {TAG. Theoretical and Applied Genetics. Theoretische und angewandte Genetik},
 doi = {10.1007/s00122-016-2675-5},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/26883048},
 howpublished = {refereed}
}

KEY MESSAGE Models based on additive marker effects and on epistatic interactions can be translated into genomic relationship models. This equivalence allows to perform predictions based on complex gene interaction models and reduces computational effort significantly. In the theory of genome-assisted prediction, the equivalence of a linear model based on independent and identically normally distributed marker effects and a model based on multivariate Gaussian distributed breeding values with genomic relationship as covariance matrix is well known. In this work, we demonstrate equivalences of marker effect models incorporating epistatic interactions and corresponding mixed models based on relationship matrices and show how to exploit these equivalences computationally for genome-assisted prediction. In particular, we show how models with epistatic interactions of higher order (e.g., three-factor interactions) translate into linear models with certain covariance matrices and demonstrate how to construct epistatic relationship matrices for the linear mixed model, if we restrict the model to interactions defined a priori. We illustrate the practical relevance of our results with a publicly available data set on grain yield of wheat lines growing in four different environments. For this purpose, we select important interactions in one environment and use this knowledge on the network of interactions to increase predictive ability of grain yield under other environmental conditions. Our results provide a guide for building relationship matrices based on knowledge on the structure of trait-related gene networks.

@article{Schwarzer2016Behavioral,
 author = {Schwarzer, Angela and Bergmann, Shana and Manitz, Juliane and K{\"u}chenhoff, Helmut and Erhard, Michael and Rauch, Elke},
 year = {2016},
 title = {Behavioral studies on the use of open water basins by {American} mink (Neovison vison)},
 url = {http://dx.doi.org/10.1016/j.jveb.2016.02.007},
 keywords = {postdoc;stat},
 pages = {19--26},
 volume = {13},
 issn = {15587878},
 journal = {Journal of Veterinary Behavior: Clinical Applications and Research},
 doi = {10.1016/j.jveb.2016.02.007},
 howpublished = {refereed}
}

@article{Tahden2016Epidemiological,
 abstract = {In 2011, a large outbreak of entero-hemorrhagic E. coli (EHEC) and hemolytic uremic syndrome (HUS) occurred in Germany. The City of Hamburg was the first focus of the epidemic and had the highest incidences among all 16 Federal States of Germany. In this article, we present epidemiological characteristics of the Hamburg notification data. Evaluating the epicurves retrospectively, we found that the first epidemiological signal of the outbreak, which was in form of a HUS case cluster, was received by local health authorities when already 99 EHEC and 48 HUS patients had experienced their first symptoms. However, only two EHEC and seven HUS patients had been notified. Middle-aged women had the highest risk for contracting the infection in Hamburg. Furthermore, we studied timeliness of case notification in the course of the outbreak. To analyze the spatial distribution of EHEC/HUS incidences in 100 districts of Hamburg, we mapped cases' residential addresses using geographic information software. We then conducted an ecological study in order to find a statistical model identifying associations between local socio-economic factors and EHEC/HUS incidences in the epidemic. We employed a Bayesian Poisson model with covariates characterizing the Hamburg districts as well as incorporating structured and unstructured spatial effects. The Deviance Information Criterion was used for stepwise variable selection. We applied different modeling approaches by using primary data, transformed data, and preselected subsets of transformed data in order to identify socio-economic factors characterizing districts where EHEC/HUS outbreak cases had their residence.},
 author = {Tahden, Maike and Manitz, Juliane and Baumgardt, Klaus and Fell, Gerhard and Kneib, Thomas and Hegasy, Guido},
 year = {2016},
 title = {Epidemiological and Ecological Characterization of the {EHEC O104:H4} Outbreak in {Hamburg, Germany}, 2011},
 url = {http://dx.doi.org/10.1371/journal.pone.0164508},
 keywords = {gen;postdoc},
 pages = {e0164508+},
 volume = {11},
 number = {10},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0164508},
 howpublished = {refereed}
}

In 2011, a large outbreak of entero-hemorrhagic E. coli (EHEC) and hemolytic uremic syndrome (HUS) occurred in Germany. The City of Hamburg was the first focus of the epidemic and had the highest incidences among all 16 Federal States of Germany. In this article, we present epidemiological characteristics of the Hamburg notification data. Evaluating the epicurves retrospectively, we found that the first epidemiological signal of the outbreak, which was in form of a HUS case cluster, was received by local health authorities when already 99 EHEC and 48 HUS patients had experienced their first symptoms. However, only two EHEC and seven HUS patients had been notified. Middle-aged women had the highest risk for contracting the infection in Hamburg. Furthermore, we studied timeliness of case notification in the course of the outbreak. To analyze the spatial distribution of EHEC/HUS incidences in 100 districts of Hamburg, we mapped cases' residential addresses using geographic information software. We then conducted an ecological study in order to find a statistical model identifying associations between local socio-economic factors and EHEC/HUS incidences in the epidemic. We employed a Bayesian Poisson model with covariates characterizing the Hamburg districts as well as incorporating structured and unstructured spatial effects. The Deviance Information Criterion was used for stepwise variable selection. We applied different modeling approaches by using primary data, transformed data, and preselected subsets of transformed data in order to identify socio-economic factors characterizing districts where EHEC/HUS outbreak cases had their residence.

@article{Wood2015Smoothing,
 abstract = {This paper discusses a general framework for smoothing parameter estimation

for models with regular likelihoods constructed in terms of unknown smooth

functions of covariates. Gaussian random effects and parametric terms may also

be present. By construction the method is numerically stable and convergent,

and enables smoothing parameter uncertainty to be quantified. The latter

enables us to fix a well known problem with AIC for such models. The smooth

functions are represented by reduced rank spline like smoothers, with

associated quadratic penalties measuring function smoothness. Model estimation

is by penalized likelihood maximization, where the smoothing parameters

controlling the extent of penalization are estimated by Laplace approximate

marginal likelihood. The methods cover, for example, generalized additive

models for non-exponential family responses (for example beta, ordered

categorical, scaled t distribution, negative binomial and Tweedie

distributions), generalized additive models for location scale and shape (for

example two stage zero inflation models, and Gaussian location-scale models),

Cox proportional hazards models and multivariate additive models. The framework

reduces the implementation of new model classes to the coding of some standard

derivatives of the log likelihood.},
 author = {Wood, Simon N. and Pya, Natalya and S{\"a}fken, Benjamin},
 year = {2016},
 title = {Smoothing Parameter and Model Selection for General Smooth Models},
 url = {http://dx.doi.org/10.1080/01621459.2016.1180986},
 keywords = {phd;stat},
 pages = {1548--1563},
 volume = {111},
 number = {516},
 issn = {0162-1459},
 journal = {Journal of the American Statistical Association},
 doi = {10.1080/01621459.2016.1180986},
 howpublished = {refereed}
}

This paper discusses a general framework for smoothing parameter estimation for models with regular likelihoods constructed in terms of unknown smooth functions of covariates. Gaussian random effects and parametric terms may also be present. By construction the method is numerically stable and convergent, and enables smoothing parameter uncertainty to be quantified. The latter enables us to fix a well known problem with AIC for such models. The smooth functions are represented by reduced rank spline like smoothers, with associated quadratic penalties measuring function smoothness. Model estimation is by penalized likelihood maximization, where the smoothing parameters controlling the extent of penalization are estimated by Laplace approximate marginal likelihood. The methods cover, for example, generalized additive models for non-exponential family responses (for example beta, ordered categorical, scaled t distribution, negative binomial and Tweedie distributions), generalized additive models for location scale and shape (for example two stage zero inflation models, and Gaussian location-scale models), Cox proportional hazards models and multivariate additive models. The framework reduces the implementation of new model classes to the coding of some standard derivatives of the log likelihood.

@article{Beduschi2015Using,
 author = {Beduschi, Tatiane and Tscharntke, Teja and Scherber, Christoph},
 year = {2015},
 title = {Using multi-level generalized path analysis to understand herbivore and parasitoid dynamics in changing landscapes},
 url = {http://dx.doi.org/10.1007/s10980-015-0224-2},
 keywords = {ecol;phd},
 pages = {1975--1986},
 volume = {30},
 number = {10},
 issn = {1572-9761},
 journal = {Landscape Ecology},
 doi = {10.1007/s10980-015-0224-2},
 howpublished = {refereed}
}

@article{Berger2015AScale,
 abstract = {The understanding of non-random association between loci, termed linkage disequilibrium (LD), plays a central role in genomic research. Since causal mutations are generally not included in genomic marker data, LD between those and available markers is essential for capturing the effects of causal loci on localizing genes responsible for traits. Thus, the interpretation of association studies requires a detailed knowledge of LD patterns. It is well known that most LD measures depend on minor allele frequencies (MAF) of the considered loci and the magnitude of LD is influenced by the physical distances between loci. In the present study, a procedure to compare the LD structure between genomic regions comprising several markers each is suggested. The approach accounts for different scaling factors, namely the distribution of MAF, the distribution of pair-wise differences in MAF, and the physical extent of compared regions, reflected by the distribution of pair-wise physical distances. In the first step, genomic regions are matched based on similarity in these scaling factors. In the second step, chromosome- and genome-wide significance tests for differences in medians of LD measures in each pair are performed. The proposed framework was applied to test the hypothesis that the average LD is different in genic and non-genic regions. This was tested with a genome-wide approach with data sets for humans (Homo sapiens), a highly selected chicken line (Gallus gallus domesticus) and the model plant Arabidopsis thaliana. In all three data sets we found a significantly higher level of LD in genic regions compared to non-genic regions. About 31{\%} more LD was detected genome-wide in genic compared to non-genic regions in Arabidopsis thaliana, followed by 13.6{\%} in human and 6{\%} chicken. Chromosome-wide comparison discovered significant differences on all 5 chromosomes in Arabidopsis thaliana and on one third of the human and of the chicken chromosomes.},
 author = {Berger, Swetlana and Schlather, Martin and de {los Campos}, Gustavo and Weigend, Steffen and Preisinger, Rudolf and Erbe, Malena and Simianer, Henner},
 year = {2015},
 title = {A Scale-Corrected Comparison of Linkage Disequilibrium Levels between Genic and Non-Genic Regions},
 url = {http://dx.doi.org/10.1371/journal.pone.0141216},
 keywords = {gen;phd},
 pages = {e0141216+},
 volume = {10},
 number = {10},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0141216},
 howpublished = {refereed}
}

The understanding of non-random association between loci, termed linkage disequilibrium (LD), plays a central role in genomic research. Since causal mutations are generally not included in genomic marker data, LD between those and available markers is essential for capturing the effects of causal loci on localizing genes responsible for traits. Thus, the interpretation of association studies requires a detailed knowledge of LD patterns. It is well known that most LD measures depend on minor allele frequencies (MAF) of the considered loci and the magnitude of LD is influenced by the physical distances between loci. In the present study, a procedure to compare the LD structure between genomic regions comprising several markers each is suggested. The approach accounts for different scaling factors, namely the distribution of MAF, the distribution of pair-wise differences in MAF, and the physical extent of compared regions, reflected by the distribution of pair-wise physical distances. In the first step, genomic regions are matched based on similarity in these scaling factors. In the second step, chromosome- and genome-wide significance tests for differences in medians of LD measures in each pair are performed. The proposed framework was applied to test the hypothesis that the average LD is different in genic and non-genic regions. This was tested with a genome-wide approach with data sets for humans (Homo sapiens), a highly selected chicken line (Gallus gallus domesticus) and the model plant Arabidopsis thaliana. In all three data sets we found a significantly higher level of LD in genic regions compared to non-genic regions. About 31% more LD was detected genome-wide in genic compared to non-genic regions in Arabidopsis thaliana, followed by 13.6% in human and 6% chicken. Chromosome-wide comparison discovered significant differences on all 5 chromosomes in Arabidopsis thaliana and on one third of the human and of the chicken chromosomes.

@article{Berger2015Effectiveness,
 abstract = {Genome-wide association studies (GWAS) have detected large numbers of variants associated with complex human traits and diseases. However, the proportion of variance explained by GWAS-significant single nucleotide polymorphisms has been usually small. This brought interest in the use of whole-genome regression (WGR) methods. However, there has been limited research on the factors that affect prediction accuracy (PA) of WGRs when applied to human data of distantly related individuals. Here, we examine, using real human genotypes and simulated phenotypes, how trait complexity, marker-quantitative trait loci (QTL) linkage disequilibrium (LD), and the model used affect the performance of WGRs. Our results indicated that the estimated rate of missing heritability is dependent on the extent of marker-QTL LD. However, this parameter was not greatly affected by trait complexity. Regarding PA our results indicated that: (a) under perfect marker-QTL LD WGR can achieve moderately high prediction accuracy, and with simple genetic architectures variable selection methods outperform shrinkage procedures and (b) under imperfect marker-QTL LD, variable selection methods can achieved reasonably good PA with simple or moderately complex genetic architectures; however, the PA of these methods deteriorated as trait complexity increases and with highly complex traits variable selection and shrinkage methods both performed poorly. This was confirmed with an analysis of human height.},
 author = {Berger, Swetlana and {P{\'e}rez-Rodrlos Campos}, Gustavo},
 year = {2015},
 title = {Effectiveness of Shrinkage and Variable Selection Methods for the Prediction of Complex Human Traits using Data from Distantly Related Individuals},
 url = {http://dx.doi.org/10.1111/ahg.12099},
 keywords = {gen;phd},
 pages = {122--135},
 volume = {79},
 number = {2},
 journal = {Annals of Human Genetics},
 doi = {10.1111/ahg.12099},
 howpublished = {refereed}
}

Genome-wide association studies (GWAS) have detected large numbers of variants associated with complex human traits and diseases. However, the proportion of variance explained by GWAS-significant single nucleotide polymorphisms has been usually small. This brought interest in the use of whole-genome regression (WGR) methods. However, there has been limited research on the factors that affect prediction accuracy (PA) of WGRs when applied to human data of distantly related individuals. Here, we examine, using real human genotypes and simulated phenotypes, how trait complexity, marker-quantitative trait loci (QTL) linkage disequilibrium (LD), and the model used affect the performance of WGRs. Our results indicated that the estimated rate of missing heritability is dependent on the extent of marker-QTL LD. However, this parameter was not greatly affected by trait complexity. Regarding PA our results indicated that: (a) under perfect marker-QTL LD WGR can achieve moderately high prediction accuracy, and with simple genetic architectures variable selection methods outperform shrinkage procedures and (b) under imperfect marker-QTL LD, variable selection methods can achieved reasonably good PA with simple or moderately complex genetic architectures; however, the PA of these methods deteriorated as trait complexity increases and with highly complex traits variable selection and shrinkage methods both performed poorly. This was confirmed with an analysis of human height.

@article{Engelke2014Estimation,
 abstract = {Estimation of extreme-value parameters from observations in the max-domain of

attraction (MDA) of a multivariate max-stable distribution commonly uses

aggregated data such as block maxima. Since we expect that additional

information is contained in the non-aggregated, single {\textquotedbl}large{\textquotedbl} observations, we

introduce a new approach of inference based on a multivariate

peaks-over-threshold method. We show that for any process in the MDA of the

frequently used H{\"u}sler-Reiss model or its spatial extension, the

Brown-Resnick process, suitably defined conditional increments asymptotically

follow a multivariate Gaussian distribution. This leads to computationally

efficient estimates of the H{\"u}sler-Reiss parameter matrix. Further, the

results enable parametric inference for Brown-Resnick processes. A simulation

study compares the performance of the new estimators to other commonly used

methods. As an application, we fit a non-isotropic Brown-Resnick process to the

extremes of 12 year data of daily wind speed measurements.},
 author = {Engelke, Sebastian and Malinowski, Alexander and Kabluchko, Zakhar and Schlather, Martin},
 year = {2015},
 title = {Estimation of {H}{\"u}sler-{Reiss} distributions and {Brown-Resnick} processes},
 keywords = {phd;stat},
 pages = {239--265},
 volume = {77},
 number = {1},
 issn = {13697412},
 journal = {Journal of the Royal Statistical Society: Series B (Statistical Methodology)},
 doi = {10.1111/rssb.12074},
 howpublished = {refereed}
}

Estimation of extreme-value parameters from observations in the max-domain of attraction (MDA) of a multivariate max-stable distribution commonly uses aggregated data such as block maxima. Since we expect that additional information is contained in the non-aggregated, single \textquotedbllarge\textquotedbl observations, we introduce a new approach of inference based on a multivariate peaks-over-threshold method. We show that for any process in the MDA of the frequently used Hüsler-Reiss model or its spatial extension, the Brown-Resnick process, suitably defined conditional increments asymptotically follow a multivariate Gaussian distribution. This leads to computationally efficient estimates of the Hüsler-Reiss parameter matrix. Further, the results enable parametric inference for Brown-Resnick processes. A simulation study compares the performance of the new estimators to other commonly used methods. As an application, we fit a non-isotropic Brown-Resnick process to the extremes of 12 year data of daily wind speed measurements.

@article{Esser2015Modelingmicrobial,
 author = {Esser, Daniel S. and Leveau, Johan H. J. and Meyer, Katrin M.},
 year = {2015},
 title = {Modeling microbial growth and dynamics},
 keywords = {ecol;phd},
 volume = {99},
 number = {21},
 journal = {Applied Microbiology and Technology},
 howpublished = {refereed}
}

@article{Esser2015SpatialScales,
 author = {Esser, Daniel S. and Leveau, Johan H. J. and Meyer, Katrin M. and Wiegand, Kerstin},
 year = {2015},
 title = {Spatial scales of interactions among bacteria and between bacteria and the leaf surface},
 keywords = {ecol;phd},
 volume = {91},
 number = {3},
 journal = {FEMS Microbiology Ecology},
 howpublished = {refereed}
}

@article{Ha2015Gene,
 abstract = {The metabolic adaptation of dairy cows during the transition period has been studied intensively in the last decades. However, until now, only few studies have paid attention to the genetic aspects of this process. Here, we present the results of a gene-based mapping and pathway analysis with the measurements of three key metabolites, (1) non-esterified fatty acids (NEFA), (2) beta-hydroxybutyrate (BHBA) and (3) glucose, characterizing the metabolic adaptability of dairy cows before and after calving. In contrast to the conventional single-marker approach, we identify 99 significant and biologically sensible genes associated with at least one of the considered phenotypes and thus giving evidence for a genetic basis of the metabolic adaptability. Moreover, our results strongly suggest three pathways involved in the metabolism of steroids and lipids are potential candidates for the adaptive regulation of dairy cows in their early lactation. From our perspective, a closer investigation of our findings will lead to a step forward in understanding the variability in the metabolic adaptability of dairy cows in their early lactation.},
 author = {Ha, Ngoc-Thuy and Gross, Josef Johann and {van Dorland}, Annette and Tetens, Jens and Thaller, Georg and Schlather, Martin and Bruckmaier, Rupert and Simianer, Henner},
 year = {2015},
 title = {Gene-based mapping and pathway analysis of metabolic traits in dairy cows},
 keywords = {gen;phd},
 pages = {e0122325},
 volume = {10},
 number = {3},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0122325},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/25789767},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4366076},
 howpublished = {refereed}
}

The metabolic adaptation of dairy cows during the transition period has been studied intensively in the last decades. However, until now, only few studies have paid attention to the genetic aspects of this process. Here, we present the results of a gene-based mapping and pathway analysis with the measurements of three key metabolites, (1) non-esterified fatty acids (NEFA), (2) beta-hydroxybutyrate (BHBA) and (3) glucose, characterizing the metabolic adaptability of dairy cows before and after calving. In contrast to the conventional single-marker approach, we identify 99 significant and biologically sensible genes associated with at least one of the considered phenotypes and thus giving evidence for a genetic basis of the metabolic adaptability. Moreover, our results strongly suggest three pathways involved in the metabolism of steroids and lipids are potential candidates for the adaptive regulation of dairy cows in their early lactation. From our perspective, a closer investigation of our findings will lead to a step forward in understanding the variability in the metabolic adaptability of dairy cows in their early lactation.

@article{Klaus2015Hedgerows,
 author = {Klaus, Felix and Bass, Julia and Marholt, Lisa and M{\"u}ller, Birte and Klatt, Bj{\"o}rn and Kormann, Urs},
 year = {2015},
 title = {Hedgerows Have a Barrier Effect and Channel Pollinator Movement in the Agricultural Landscape},
 url = {http://dx.doi.org/10.1515/jlecol-2015-0001},
 keywords = {ecol;phd},
 volume = {8},
 number = {1},
 issn = {1805-4196},
 journal = {Journal of Landscape Ecology},
 doi = {10.1515/jlecol-2015-0001},
 howpublished = {refereed}
}

@article{Klein2015Bayesian,
 author = {Klein, Nadja and Kneib, Thomas and Klasen, Stephan and Lang, Stefan},
 year = {2015},
 title = {Bayesian structured additive distributional regression for multivariate responses},
 url = {http://dx.doi.org/10.1111/rssc.12090},
 keywords = {postdoc;stat},
 pages = {569--591},
 volume = {64},
 number = {4},
 issn = {00359254},
 journal = {Journal of the Royal Statistical Society: Series C (Applied Statistics)},
 doi = {10.1111/rssc.12090},
 howpublished = {refereed}
}

@article{Klein2015BayesianGeneralized,
 abstract = {Frequent problems in applied research preventing the application of the classical Poisson log-linear model for analysing count data include overdispersion, an excess of zeros compared to the Poisson distribution, correlated responses, as well as complex predictor structures comprising nonlinear effects of continuous covariates, interactions or spatial effects. We propose a general class of Bayesian generalised additive models for zero-inflated and overdispersed count data within the framework of generalised additive models for location, scale, and shape where semiparametric predictors can be specified for several parameters of a count data distribution. As standard options for applied work we consider the zero-inflated Poisson, the negative binomial and the zero-inflated negative binomial distribution. The additive predictor specifications rely on basis function approximations for the different types of effects in combination with Gaussian smoothness priors. We develop Bayesian inference based on Markov chain Monte Carlo simulation techniques where suitable proposal densities are constructed based on iteratively weighted least squares approximations to the full conditionals. To ensure practicability of the inference, we consider theoretical properties like the involved question whether the joint posterior is proper. The proposed approach is evaluated in simulation studies and applied to count data arising from patent citations and claim frequencies in car insurances. For the comparison of models with respect to the distribution, we consider quantile residuals as an effective graphical device and scoring rules that allow to quantify the predictive ability of the models. The deviance information criterion is used to select appropriate predictor specifications once a response distribution has been chosen.},
 author = {Klein, Nadja and Kneib, Thomas and Lang, Stefan},
 year = {2015},
 title = {Bayesian Generalized Additive Models for Location, Scale, and Shape for Zero-Inflated and Overdispersed~Count~Data},
 url = {http://dx.doi.org/10.1080/01621459.2014.912955},
 keywords = {phd;stat},
 pages = {405--419},
 volume = {110},
 number = {509},
 issn = {0162-1459},
 journal = {Journal of the American Statistical Association},
 doi = {10.1080/01621459.2014.912955},
 howpublished = {refereed}
}

Frequent problems in applied research preventing the application of the classical Poisson log-linear model for analysing count data include overdispersion, an excess of zeros compared to the Poisson distribution, correlated responses, as well as complex predictor structures comprising nonlinear effects of continuous covariates, interactions or spatial effects. We propose a general class of Bayesian generalised additive models for zero-inflated and overdispersed count data within the framework of generalised additive models for location, scale, and shape where semiparametric predictors can be specified for several parameters of a count data distribution. As standard options for applied work we consider the zero-inflated Poisson, the negative binomial and the zero-inflated negative binomial distribution. The additive predictor specifications rely on basis function approximations for the different types of effects in combination with Gaussian smoothness priors. We develop Bayesian inference based on Markov chain Monte Carlo simulation techniques where suitable proposal densities are constructed based on iteratively weighted least squares approximations to the full conditionals. To ensure practicability of the inference, we consider theoretical properties like the involved question whether the joint posterior is proper. The proposed approach is evaluated in simulation studies and applied to count data arising from patent citations and claim frequencies in car insurances. For the comparison of models with respect to the distribution, we consider quantile residuals as an effective graphical device and scoring rules that allow to quantify the predictive ability of the models. The deviance information criterion is used to select appropriate predictor specifications once a response distribution has been chosen.

@article{Klein2015BayesianStructured,
 abstract = {We propose a generic Bayesian framework for inference in distributional

regression models in which each parameter of a potentially complex response

distribution and not only the mean is related to a structured additive

predictor. The latter is composed additively of a variety of different

functional effect types such as nonlinear effects, spatial effects, random

coefficients, interaction surfaces or other (possibly nonstandard) basis

function representations. To enforce specific properties of the functional

effects such as smoothness, informative multivariate Gaussian priors are

assigned to the basis function coefficients. Inference can then be based on

computationally efficient Markov chain Monte Carlo simulation techniques where

a generic procedure makes use of distribution-specific iteratively weighted

least squares approximations to the full conditionals. The framework of

distributional regression encompasses many special cases relevant for treating

nonstandard response structures such as highly skewed nonnegative responses,

overdispersed and zero-inflated counts or shares including the possibility for

zero- and one-inflation. We discuss distributional regression along a study on

determinants of labour incomes for full-time working males in Germany with a

particular focus on regional differences after the German reunification.

Controlling for age, education, work experience and local disparities, we

estimate full conditional income distributions allowing us to study various

distributional quantities such as moments, quantiles or inequality measures in

a consistent manner in one joint model. Detailed guidance on practical aspects

of model choice including the selection of several competing distributions for

labour incomes and the consideration of different covariate effects on the

income distribution complete the distributional regression analysis. We find

that next to a lower expected income, full-time working men in East Germany

also face a more unequal income distribution than men in the West, ceteris

paribus.},
 author = {Klein, Nadja and Kneib, Thomas and Lang, Stefan and Sohn, Alexander},
 year = {2015},
 title = {Bayesian structured additive distributional regression with an application to regional income inequality in {Germany}},
 url = {http://dx.doi.org/10.1214/15-aoas823},
 keywords = {phd;stat},
 pages = {1024--1052},
 volume = {9},
 number = {2},
 issn = {1932-6157},
 journal = {The Annals of Applied Statistics},
 doi = {10.1214/15-aoas823},
 howpublished = {refereed}
}

We propose a generic Bayesian framework for inference in distributional regression models in which each parameter of a potentially complex response distribution and not only the mean is related to a structured additive predictor. The latter is composed additively of a variety of different functional effect types such as nonlinear effects, spatial effects, random coefficients, interaction surfaces or other (possibly nonstandard) basis function representations. To enforce specific properties of the functional effects such as smoothness, informative multivariate Gaussian priors are assigned to the basis function coefficients. Inference can then be based on computationally efficient Markov chain Monte Carlo simulation techniques where a generic procedure makes use of distribution-specific iteratively weighted least squares approximations to the full conditionals. The framework of distributional regression encompasses many special cases relevant for treating nonstandard response structures such as highly skewed nonnegative responses, overdispersed and zero-inflated counts or shares including the possibility for zero- and one-inflation. We discuss distributional regression along a study on determinants of labour incomes for full-time working males in Germany with a particular focus on regional differences after the German reunification. Controlling for age, education, work experience and local disparities, we estimate full conditional income distributions allowing us to study various distributional quantities such as moments, quantiles or inequality measures in a consistent manner in one joint model. Detailed guidance on practical aspects of model choice including the selection of several competing distributions for labour incomes and the consideration of different covariate effects on the income distribution complete the distributional regression analysis. We find that next to a lower expected income, full-time working men in East Germany also face a more unequal income distribution than men in the West, ceteris paribus.

@article{Klein2015Simultaneous,
 author = {Klein, Nadja and Kneib, Thomas},
 year = {2015},
 title = {Simultaneous inference in structured additive conditional copula regression models: a unifying {Bayesian} approach},
 url = {http://dx.doi.org/10.1007/s11222-015-9573-6},
 keywords = {phd;stat},
 pages = {841--860},
 volume = {26},
 number = {4},
 issn = {0960-3174},
 journal = {Statistics and Computing},
 doi = {10.1007/s11222-015-9573-6},
 howpublished = {refereed}
}

@article{Kormann2015Local,
 abstract = {Aim Biodiversity across the globe is heavily eroded by intensified management at local and landscape scales. Species communities of calcareous grasslands, which are among Europe's most diverse habitats, are severely threatened by the cessation of appropriate traditional management, loss of habitat connectivity and simplification of the surrounding landscape. However, our understanding of these often interrelated factors remains limited, in particular for trait-mediated responses across taxa. Here, we test the independent effects of local management (grazing, mowing and abandonment), habitat connectivity (measured by a connectivity index) and landscape complexity (indicated by the percentage of arable land) on nine taxa: plants, butterflies, bees, grasshoppers, hoverflies, spiders, true bugs, rove beetles and leafhoppers on small semi-natural calcareous grassland remnants ({\textless}~1~ha). Location Central Germany. Methods We use a joint analysis across taxa to identify general and trait-mediated responses (body size and Red List status) in species richness, abundance and community composition. Results We identified three key drivers of local diversity patterns: First, an increasing proportion of arable land from 10{\%} to 80{\%} led to a 29{\%} loss of overall species richness. Second, despite differences between taxa, increasing habitat connectivity generally enhanced species richness. Connectivity effects were more accentuated in the large species per taxon, which can be expected to be good dispersers. Finally, grazing reduced species richness and abundance much more than annual mowing or short-term abandonment (5--15~years), in particular for red-listed species. We attribute this to plant resource removal through overgrazing and trampling. Main conclusions For the conservation management of small calcareous grasslands, we advocate an alternating strategy of mowing or lenient grazing and short-term abandonment, prioritizing connected fragments surrounded by diverse landscapes. Despite taxon-specific responses, our study across nine taxa demonstrates universal, trait-mediated effects of management, landscape complexity and connectivity on local biodiversity in fragmented communities.},
 author = {Kormann, Urs and R{\"o}sch, Verena and Bat{\'a}ry, P{\'e}ter and Tscharntke, Teja and Orci, Kirill M. and Samu, Ferenc and Scherber, Christoph},
 year = {2015},
 title = {Local and landscape management drive trait-mediated biodiversity of nine taxa on small grassland fragments},
 url = {http://dx.doi.org/10.1111/ddi.12324},
 keywords = {ecol;phd},
 pages = {1204--1217},
 volume = {21},
 number = {10},
 issn = {13669516},
 journal = {Diversity and Distributions},
 doi = {10.1111/ddi.12324},
 howpublished = {refereed}
}

Aim Biodiversity across the globe is heavily eroded by intensified management at local and landscape scales. Species communities of calcareous grasslands, which are among Europe's most diverse habitats, are severely threatened by the cessation of appropriate traditional management, loss of habitat connectivity and simplification of the surrounding landscape. However, our understanding of these often interrelated factors remains limited, in particular for trait-mediated responses across taxa. Here, we test the independent effects of local management (grazing, mowing and abandonment), habitat connectivity (measured by a connectivity index) and landscape complexity (indicated by the percentage of arable land) on nine taxa: plants, butterflies, bees, grasshoppers, hoverflies, spiders, true bugs, rove beetles and leafhoppers on small semi-natural calcareous grassland remnants (\textless 1 ha). Location Central Germany. Methods We use a joint analysis across taxa to identify general and trait-mediated responses (body size and Red List status) in species richness, abundance and community composition. Results We identified three key drivers of local diversity patterns: First, an increasing proportion of arable land from 10% to 80% led to a 29% loss of overall species richness. Second, despite differences between taxa, increasing habitat connectivity generally enhanced species richness. Connectivity effects were more accentuated in the large species per taxon, which can be expected to be good dispersers. Finally, grazing reduced species richness and abundance much more than annual mowing or short-term abandonment (5–15 years), in particular for red-listed species. We attribute this to plant resource removal through overgrazing and trampling. Main conclusions For the conservation management of small calcareous grasslands, we advocate an alternating strategy of mowing or lenient grazing and short-term abandonment, prioritizing connected fragments surrounded by diverse landscapes. Despite taxon-specific responses, our study across nine taxa demonstrates universal, trait-mediated effects of management, landscape complexity and connectivity on local biodiversity in fragmented communities.

@article{Ober2015Accounting,
 abstract = {The ability to predict quantitative trait phenotypes from molecular polymorphism data will revolutionize evolutionary biology, medicine and human biology, and animal and plant breeding. Efforts to map quantitative trait loci have yielded novel insights into the biology of quantitative traits, but the combination of individually significant quantitative trait loci typically has low predictive ability. Utilizing all segregating variants can give good predictive ability in plant and animal breeding populations, but gives little insight into trait biology. Here, we used the Drosophila Genetic Reference Panel to perform both a genome wide association analysis and genomic prediction for the fitness-related trait chill coma recovery time. We found substantial total genetic variation for chill coma recovery time, with a genetic architecture that differs between males and females, a small number of molecular variants with large main effects, and evidence for epistasis. Although the top additive variants explained 36{\%} (17{\%}) of the genetic variance among lines in females (males), the predictive ability using genomic best linear unbiased prediction and a relationship matrix using all common segregating variants was very low for females and zero for males. We hypothesized that the low predictive ability was due to the mismatch between the infinitesimal genetic architecture assumed by the genomic best linear unbiased prediction model and the true genetic architecture of chill coma recovery time. Indeed, we found that the predictive ability of the genomic best linear unbiased prediction model is markedly improved when we combine quantitative trait locus mapping with genomic prediction by only including the top variants associated with main and epistatic effects in the relationship matrix. This trait-associated prediction approach has the advantage that it yields biologically interpretable prediction models.},
 author = {Ober, Ulrike and Huang, Wen and Magwire, Michael and Schlather, Martin and Simianer, Henner and Mackay, Trudy F. C.},
 year = {2015},
 title = {Accounting for genetic architecture improves sequence based genomic prediction for a \textit{Drosophila} fitness trait},
 keywords = {gen;postdoc},
 pages = {e0126880},
 volume = {10},
 number = {5},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0126880},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/25950439},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4423967},
 howpublished = {refereed}
}

The ability to predict quantitative trait phenotypes from molecular polymorphism data will revolutionize evolutionary biology, medicine and human biology, and animal and plant breeding. Efforts to map quantitative trait loci have yielded novel insights into the biology of quantitative traits, but the combination of individually significant quantitative trait loci typically has low predictive ability. Utilizing all segregating variants can give good predictive ability in plant and animal breeding populations, but gives little insight into trait biology. Here, we used the Drosophila Genetic Reference Panel to perform both a genome wide association analysis and genomic prediction for the fitness-related trait chill coma recovery time. We found substantial total genetic variation for chill coma recovery time, with a genetic architecture that differs between males and females, a small number of molecular variants with large main effects, and evidence for epistasis. Although the top additive variants explained 36% (17%) of the genetic variance among lines in females (males), the predictive ability using genomic best linear unbiased prediction and a relationship matrix using all common segregating variants was very low for females and zero for males. We hypothesized that the low predictive ability was due to the mismatch between the infinitesimal genetic architecture assumed by the genomic best linear unbiased prediction model and the true genetic architecture of chill coma recovery time. Indeed, we found that the predictive ability of the genomic best linear unbiased prediction model is markedly improved when we combine quantitative trait locus mapping with genomic prediction by only including the top variants associated with main and epistatic effects in the relationship matrix. This trait-associated prediction approach has the advantage that it yields biologically interpretable prediction models.

@incollection{Razen2015Hedonic,
 author = {Razen, Alexander and Brunauer, Wolfgang and Klein, Nadja and Lang, Stefan and Umlauf, Nikolaus},
 title = {Hedonic House Price Modeling Based on Multilevel Structured Additive Regression},
 url = {http://dx.doi.org/10.1007/978-3-319-11469-9\textunderscore 5},
 keywords = {phd;stat},
 pages = {97--122},
 publisher = {{Springer International Publishing}},
 isbn = {978-3-319-11468-2},
 editor = {Helbich, Marco and {Jokar Arsanjani}, Jamal and Leitner, Michael},
 booktitle = {Computational Approaches for Urban Environments},
 year = {2015},
 address = {Cham},
 doi = {10.1007/978-3-319-11469-9{\textunderscore }5},
 howpublished = {refereed}
}

@article{Rosenberger2015Combining,
 abstract = {Gene-set analysis (GSA) methods are used as complementary approaches to genome-wide association studies (GWASs). The single marker association estimates of a predefined set of genes are either contrasted with those of all remaining genes or with a null non-associated background. To pool the p-values from several GSAs, it is important to take into account the concordance of the observed patterns resulting from single marker association point estimates across any given gene set. Here we propose an enhanced version of Fisher's inverse \textgreek{q}2-method META-GSA, however weighting each study to account for imperfect correlation between association patterns. We investigated the performance of META-GSA by simulating GWASs with 500 cases and 500 controls at 100 diallelic markers in 20 different scenarios, simulating different relative risks between 1 and 1.5 in gene sets of 10 genes. Wilcoxon's rank sum test was applied as GSA for each study. We found that META-GSA has greater power to discover truly associated gene sets than simple pooling of the p-values, by e.g. 59{\%} versus 37{\%}, when the true relative risk for 5 of 10 genes was assume to be 1.5. Under the null hypothesis of no difference in the true association pattern between the gene set of interest and the set of remaining genes, the results of both approaches are almost uncorrelated. We recommend not relying on p-values alone when combining the results of independent GSAs. We applied META-GSA to pool the results of four case-control GWASs of lung cancer risk (Central European Study and Toronto/Lunenfeld-Tanenbaum Research Institute Study; German Lung Cancer Study and MD Anderson Cancer Center Study), which had already been analyzed separately with four different GSA methods (EASE; SLAT, mSUMSTAT and GenGen). This application revealed the pathway GO0015291  '' transmembrane transporter activity'' as significantly enriched with associated genes (GSA-method: EASE, p = 0.0315 corrected for multiple testing). Similar results were found for GO0015464  '' acetylcholine receptor activity'' but only when not corrected for multiple testing (all GSA-methods applied; p$\approx$0.02).},
 author = {Rosenberger, Albert and Friedrichs, Stefanie and {[...]} and Bickeb{\"o}ller, Heike},
 year = {2015},
 title = {{META-GSA}: Combining Findings from Gene-Set Analyses across Several Genome-Wide Association Studies},
 url = {http://dx.doi.org/10.1371/journal.pone.0140179},
 keywords = {gen;phd},
 pages = {e0140179+},
 volume = {10},
 number = {10},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0140179},
 howpublished = {refereed}
}

Gene-set analysis (GSA) methods are used as complementary approaches to genome-wide association studies (GWASs). The single marker association estimates of a predefined set of genes are either contrasted with those of all remaining genes or with a null non-associated background. To pool the p-values from several GSAs, it is important to take into account the concordance of the observed patterns resulting from single marker association point estimates across any given gene set. Here we propose an enhanced version of Fisher's inverse \textgreekq2-method META-GSA, however weighting each study to account for imperfect correlation between association patterns. We investigated the performance of META-GSA by simulating GWASs with 500 cases and 500 controls at 100 diallelic markers in 20 different scenarios, simulating different relative risks between 1 and 1.5 in gene sets of 10 genes. Wilcoxon's rank sum test was applied as GSA for each study. We found that META-GSA has greater power to discover truly associated gene sets than simple pooling of the p-values, by e.g. 59% versus 37%, when the true relative risk for 5 of 10 genes was assume to be 1.5. Under the null hypothesis of no difference in the true association pattern between the gene set of interest and the set of remaining genes, the results of both approaches are almost uncorrelated. We recommend not relying on p-values alone when combining the results of independent GSAs. We applied META-GSA to pool the results of four case-control GWASs of lung cancer risk (Central European Study and Toronto/Lunenfeld-Tanenbaum Research Institute Study; German Lung Cancer Study and MD Anderson Cancer Center Study), which had already been analyzed separately with four different GSA methods (EASE; SLAT, mSUMSTAT and GenGen). This application revealed the pathway GO0015291 '' transmembrane transporter activity'' as significantly enriched with associated genes (GSA-method: EASE, p = 0.0315 corrected for multiple testing). Similar results were found for GO0015464 '' acetylcholine receptor activity'' but only when not corrected for multiple testing (all GSA-methods applied; p$≈$0.02).

@article{Sohn2015ASemiparametric,
 author = {Sohn, Alexander and Klein, Nadja and Kneib, Thomas},
 year = {2015},
 title = {A Semiparametric Analysis of Conditional Income Distributions},
 url = {http://dx.doi.org/10.3790/schm.135.1.13},
 keywords = {econ;phd;stat},
 pages = {13--22},
 volume = {135},
 number = {1},
 issn = {2568-7603},
 journal = {Journal of Contextual Economics (Schmollers Jahrbuch)},
 doi = {10.3790/schm.135.1.13},
 howpublished = {refereed}
}

@article{Sutcliffe2015Harnessing,
 author = {Sutcliffe, Laura M. E. and Bat{\'a}ry, P{\'e}ter and Kormann, Urs and {[...]} and Tscharntke, Teja},
 year = {2015},
 title = {Harnessing the biodiversity value of {Central} and {Eastern European} farmland},
 url = {http://dx.doi.org/10.1111/ddi.12288},
 keywords = {ecol;phd},
 pages = {722--730},
 volume = {21},
 number = {6},
 issn = {13669516},
 journal = {Diversity and Distributions},
 doi = {10.1111/ddi.12288},
 howpublished = {refereed}
}

@article{Zahawi2015Sixty,
 author = {Zahawi, Rakan A. and Duran, Guillermo and Kormann, Urs},
 year = {2015},
 title = {Sixty-Seven Years of Land-Use Change in {Southern Costa Rica}},
 url = {http://dx.doi.org/10.1371/journal.pone.0143554},
 keywords = {ecol;phd},
 pages = {e0143554+},
 volume = {10},
 number = {11},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0143554},
 howpublished = {refereed}
}

@article{Zhang2015Accuracy,
 author = {Zhang, Zhe and Erbe, Malena and He, Jinlong and Ober, Ulrike and Gao, Ning and Zhang, Hao and Simianer, Henner and Liand, Jiaqi},
 year = {2015},
 title = {Accuracy of Whole-Genome Prediction Using a Genetic Architecture-Enhanced Variance-Covariance Matrix},
 keywords = {gen;phd},
 volume = {g3.114.016261},
 journal = {G3: Genes|Genomes|Genetics},
 howpublished = {refereed}
}

@article{Chisholm2014Temporal,
 abstract = {Long-term surveys of entire communities of species are needed to measure fluctuations in natural populations and elucidate the mechanisms driving population dynamics and community assembly. We analysed changes in abundance of over 4000 tree species in 12 forests across the world over periods of 6--28~years. Abundance fluctuations in all forests are large and consistent with population dynamics models in which temporal environmental variance plays a central role. At some sites we identify clear environmental drivers, such as fire and drought, that could underlie these patterns, but at other sites there is a need for further research to identify drivers. In addition, cross-site comparisons showed that abundance fluctuations were smaller at species-rich sites, consistent with the idea that stable environmental conditions promote higher diversity. Much community ecology theory emphasises demographic variance and niche stabilisation; we encourage the development of theory in which temporal environmental variance plays a central role.},
 author = {Chisholm, Ryan A. and Condit, Richard and Rahman and Baker, Patrick J. and Bunyavejchewin, Sarayudh and Chen, Yu-Yun and Chuyong, George and Dattaraja, H. S. and Davies, Stuart and Ewango, Corneille E. N. and Gunatilleke, C. V. S. and {Nimal Gunatilleke, I. A. U.} and Hubbell, Stephen and Kenfack, David and Kiratiprayoon, Somboon and Lin, Yiching and Makana, Jean-Remy and Pongpattananurak, Nantachai and Pulla, Sandeep and Punchi-Manage, Ruwan and Sukumar, Raman and Su, Sheng-Hsin and Sun, I-Fang and Suresh, H. S. and Tan, Sylvester and Thomas, Duncan and Yap, Sandra},
 year = {2014},
 title = {Temporal variability of forest communities: empirical estimates of population change in 4000 tree species},
 url = {http://dx.doi.org/10.1111/ele.12296},
 keywords = {ecol;phd},
 pages = {855--865},
 volume = {17},
 number = {7},
 issn = {1461023X},
 journal = {Ecology Letters},
 doi = {10.1111/ele.12296},
 howpublished = {refereed}
}

Long-term surveys of entire communities of species are needed to measure fluctuations in natural populations and elucidate the mechanisms driving population dynamics and community assembly. We analysed changes in abundance of over 4000 tree species in 12 forests across the world over periods of 6–28 years. Abundance fluctuations in all forests are large and consistent with population dynamics models in which temporal environmental variance plays a central role. At some sites we identify clear environmental drivers, such as fire and drought, that could underlie these patterns, but at other sites there is a need for further research to identify drivers. In addition, cross-site comparisons showed that abundance fluctuations were smaller at species-rich sites, consistent with the idea that stable environmental conditions promote higher diversity. Much community ecology theory emphasises demographic variance and niche stabilisation; we encourage the development of theory in which temporal environmental variance plays a central role.

@article{Engelke2014Representations,
 abstract = {This paper provides the basis for new methods of inference for max-stable

processes $\backslash$xi$\backslash$ on general spaces that admit a certain incremental

representation, which, in important cases, has a much simpler structure than

the max-stable process itself. A corresponding peaks-over-threshold approach

will incorporate all single events that are extreme in some sense and will

therefore rely on a substantially larger amount of data in comparison to

estimation procedures based on block maxima. Conditioning a process $\backslash$eta$\backslash$ in

the max-domain of attraction of $\backslash$xi$\backslash$ on being extremal, several convergence

results for the increments of $\backslash$eta$\backslash$ are proved. In a similar way, the shape

functions of mixed moving maxima (M3) processes can be extracted from suitably

conditioned single events $\backslash$eta. Connecting the two approaches, transformation

formulae for processes that admit both an incremental and an M3 representation

are identified.},
 author = {Engelke, Sebastian and Malinowski, Alexander and Oesting, Marco and Schlather, Martin},
 year = {2014},
 title = {Statistical Inference for Max-Stable Processes by Conditioning on Extreme Events},
 keywords = {phd;stat},
 pages = {478--495},
 volume = {46},
 number = {02},
 issn = {0001-8678},
 journal = {Advances in Applied Probability},
 doi = {10.1239/aap/1401369703},
 howpublished = {refereed}
}

This paper provides the basis for new methods of inference for max-stable processes $\$xi$\$ on general spaces that admit a certain incremental representation, which, in important cases, has a much simpler structure than the max-stable process itself. A corresponding peaks-over-threshold approach will incorporate all single events that are extreme in some sense and will therefore rely on a substantially larger amount of data in comparison to estimation procedures based on block maxima. Conditioning a process $\$eta$\$ in the max-domain of attraction of $\$xi$\$ on being extremal, several convergence results for the increments of $\$eta$\$ are proved. In a similar way, the shape functions of mixed moving maxima (M3) processes can be extracted from suitably conditioned single events $\$eta. Connecting the two approaches, transformation formulae for processes that admit both an incremental and an M3 representation are identified.

@article{Ha2014Coverage,
 abstract = {To answer the question as to which commercial high-density SNP chip covers most of the human genome given a fixed budget, we compared the performance of 12 chips of different sizes released by Affymetrix and Illumina for the European, Asian, and African populations. These include Affymetrix' relatively new population-optimized arrays, whose SNP sets are each tailored toward a specific ethnicity. Our evaluation of the chips included the use of two measures, efficiency and cost-benefit ratio, which we developed as supplements to genetic coverage. Unlike coverage, these measures factor in the price of a chip or its substitute size (number of SNPs on chip), allowing comparisons to be drawn between differently priced chips. In this fashion, we identified the Affymetrix population-optimized arrays as offering the most cost-effective coverage for the Asian and African population. For the European population, we established the Illumina Human Omni 2.5-8 as the preferred choice. Interestingly, the Affymetrix chip tailored toward an Eastern Asian subpopulation performed well for all three populations investigated. However, our coverage estimates calculated for all chips proved much lower than those advertised by the producers. All our analyses were based on the 1000 Genome Project as reference population.},
 author = {Ha, Ngoc-Thuy and Freytag, Saskia and Bickeb{\"o}ller, Heike},
 year = {2014},
 title = {Coverage and efficiency in current {SNP} chips},
 url = {http://dx.doi.org/10.1038/ejhg.2013.304},
 keywords = {gen;phd},
 pages = {1124--1130},
 volume = {22},
 number = {9},
 journal = {European Journal of Human Genetics},
 doi = {10.1038/ejhg.2013.304},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/24448550},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4135415},
 howpublished = {refereed}
}

To answer the question as to which commercial high-density SNP chip covers most of the human genome given a fixed budget, we compared the performance of 12 chips of different sizes released by Affymetrix and Illumina for the European, Asian, and African populations. These include Affymetrix' relatively new population-optimized arrays, whose SNP sets are each tailored toward a specific ethnicity. Our evaluation of the chips included the use of two measures, efficiency and cost-benefit ratio, which we developed as supplements to genetic coverage. Unlike coverage, these measures factor in the price of a chip or its substitute size (number of SNPs on chip), allowing comparisons to be drawn between differently priced chips. In this fashion, we identified the Affymetrix population-optimized arrays as offering the most cost-effective coverage for the Asian and African population. For the European population, we established the Illumina Human Omni 2.5-8 as the preferred choice. Interestingly, the Affymetrix chip tailored toward an Eastern Asian subpopulation performed well for all three populations investigated. However, our coverage estimates calculated for all chips proved much lower than those advertised by the producers. All our analyses were based on the 1000 Genome Project as reference population.

@article{Klein2014Nonlife,
 author = {Klein, N. and Denuit, M. and Lang, S. and Kneib, Thomas},
 year = {2014},
 title = {Nonlife Ratemaking and Risk Management with {Bayesian} Additive Models for Location, Scale and Shape},
 url = {http://dx.doi.org/10.1016/j.insmatheco.2014.02.001},
 keywords = {phd},
 pages = {225--249},
 volume = {55},
 journal = {Insurance: Mathematics and Economics},
 doi = {10.1016/j.insmatheco.2014.02.001},
 howpublished = {refereed}
}

@article{Malzahn2014Kernel,
 abstract = {The kernel score statistic is a global covariance component test over a set of genetic markers. It provides a flexible modeling framework and does not collapse marker information. We generalize the kernel score statistic to allow for familial dependencies and to adjust for random confounder effects. With this extension, we adjust our analysis of real and simulated baseline systolic blood pressure for polygenic familial background. We find that the kernel score test gains appreciably in power through the use of sequencing compared to tag-single-nucleotide polymorphisms for very rare single nucleotide polymorphisms with {\textless}1{\%} minor allele frequency.},
 author = {Malzahn, D{\"o}rthe and Friedrichs, Stefanie and Rosenberger, Albert and Bickeb{\"o}ller, Heike},
 year = {2014},
 title = {Kernel score statistic for dependent data},
 keywords = {gen;phd},
 pages = {S41},
 volume = {8},
 number = {Suppl 1},
 issn = {1753-6561},
 journal = {BMC Proceedings},
 doi = {10.1186/1753-6561-8-S1-S41},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/25519324},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4143755},
 howpublished = {refereed}
}

The kernel score statistic is a global covariance component test over a set of genetic markers. It provides a flexible modeling framework and does not collapse marker information. We generalize the kernel score statistic to allow for familial dependencies and to adjust for random confounder effects. With this extension, we adjust our analysis of real and simulated baseline systolic blood pressure for polygenic familial background. We find that the kernel score test gains appreciably in power through the use of sequencing compared to tag-single-nucleotide polymorphisms for very rare single nucleotide polymorphisms with \textless1% minor allele frequency.

@article{Manitz2014Origin,
 author = {Manitz, Juliane and Kneib, Thomas and Schlather, Martin and Helbing, Dirk and Brockmann, Dirk},
 year = {2014},
 title = {Origin Detection During Food-borne Disease Outbreaks --- {A} Case Study of the 2011 {EHEC/HUS} Outbreak in {Germany}},
 url = {http://dx.doi.org/10.1371/currents.outbreaks.f3fdeb08c5b9de7c09ed9cbcef5f01f2},
 keywords = {phd;stat},
 issn = {2157-3999},
 journal = {PLoS Currents: Outbreaks},
 doi = {10.1371/currents.outbreaks.f3fdeb08c5b9de7c09ed9cbcef5f01f2},
 howpublished = {refereed}
}

@article{Pfeifer2014Biofrag,
 abstract = {Habitat fragmentation studies have produced complex results that are challenging to synthesize. Inconsistencies among studies may result from variation in the choice of landscape metrics and response variables, which is often compounded by a lack of key statistical or methodological information. Collating primary datasets on biodiversity responses to fragmentation in a consistent and flexible database permits simple data retrieval for subsequent analyses. We present a relational database that links such field data to taxonomic nomenclature, spatial and temporal plot attributes, and environmental characteristics. Field assessments include measurements of the response(s) (e.g., presence, abundance, ground cover) of one or more species linked to plots in fragments within a partially forested landscape. The database currently holds 9830 unique species recorded in plots of 58 unique landscapes in six of eight realms: mammals 315, birds 1286, herptiles 460, insects 4521, spiders 204, other arthropods 85, gastropods 70, annelids 8, platyhelminthes 4, Onychophora 2, vascular plants 2112, nonvascular plants and lichens 320, and fungi 449. Three landscapes were sampled as long-term time series ({\textgreater}10 years). Seven hundred and eleven species are found in two or more landscapes. Consolidating the substantial amount of primary data available on biodiversity responses to fragmentation in the context of land-use change and natural disturbances is an essential part of understanding the effects of increasing anthropogenic pressures on land. The consistent format of this database facilitates testing of generalizations concerning biologic responses to fragmentation across diverse systems and taxa. It also allows the re-examination of existing datasets with alternative landscape metrics and robust statistical methods, for example, helping to address pseudo-replication problems. The database can thus help researchers in producing broad syntheses of the effects of land use. The database is dynamic and inclusive, and contributions from individual and large-scale data-collection efforts are welcome.},
 author = {Pfeifer, Marion and {[...]} and Kormann, Urs and Scherber, Christoph and Tscharntke, Teja and Tyre, Andrew J. and {Urbina Cardona}, Jose N. and Vasconcelos, Heraldo L. and Wearn, Oliver and Wells, Konstans and Willig, Michael R. and Wood, Eric and Young, Richard P. and Bradley, Andrew V. and Ewers, Robert M.},
 year = {2014},
 title = {{BIOFRAG} -- a new database for analyzing {BIOdiversity responses to forest FRAGmentation}},
 keywords = {ecol;phd},
 pages = {1524--1537},
 volume = {4},
 number = {9},
 journal = {Ecology and Evolution},
 doi = {10.1002/ece3.1036},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/24967073},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4063456},
 howpublished = {refereed}
}

Habitat fragmentation studies have produced complex results that are challenging to synthesize. Inconsistencies among studies may result from variation in the choice of landscape metrics and response variables, which is often compounded by a lack of key statistical or methodological information. Collating primary datasets on biodiversity responses to fragmentation in a consistent and flexible database permits simple data retrieval for subsequent analyses. We present a relational database that links such field data to taxonomic nomenclature, spatial and temporal plot attributes, and environmental characteristics. Field assessments include measurements of the response(s) (e.g., presence, abundance, ground cover) of one or more species linked to plots in fragments within a partially forested landscape. The database currently holds 9830 unique species recorded in plots of 58 unique landscapes in six of eight realms: mammals 315, birds 1286, herptiles 460, insects 4521, spiders 204, other arthropods 85, gastropods 70, annelids 8, platyhelminthes 4, Onychophora 2, vascular plants 2112, nonvascular plants and lichens 320, and fungi 449. Three landscapes were sampled as long-term time series (\textgreater10 years). Seven hundred and eleven species are found in two or more landscapes. Consolidating the substantial amount of primary data available on biodiversity responses to fragmentation in the context of land-use change and natural disturbances is an essential part of understanding the effects of increasing anthropogenic pressures on land. The consistent format of this database facilitates testing of generalizations concerning biologic responses to fragmentation across diverse systems and taxa. It also allows the re-examination of existing datasets with alternative landscape metrics and robust statistical methods, for example, helping to address pseudo-replication problems. The database can thus help researchers in producing broad syntheses of the effects of land use. The database is dynamic and inclusive, and contributions from individual and large-scale data-collection efforts are welcome.

@article{PunchiManage2014Effect,
 author = {Punchi-Manage, Ruwan and Wiegand, Thorsten and Wiegand, Kerstin and Getzin, Stephan and Gunatilleke, C. V. Savitri and {Gunatilleke, I. A. U. Nimal}},
 year = {2014},
 title = {Effect of spatial processes and topography on structuring species assemblages in a {Sri Lankan} dipterocarp forest},
 url = {http://dx.doi.org/10.1890/12-2102.1},
 keywords = {ecol;phd},
 pages = {376--386},
 volume = {95},
 number = {2},
 issn = {0012-9658},
 journal = {Ecology},
 doi = {10.1890/12-2102.1},
 howpublished = {refereed}
}

@article{RejouMechain2014Local,
 abstract = {Advances in forest carbon mapping have the potential to greatly reduce uncertainties in the global carbon budget and to facilitate effective emissions mitigation strategies such as REDD+. Though broad scale mapping is based primarily on remote sensing data, the accuracy of resulting forest carbon stock estimates depends critically on the quality of field measurements and calibration procedures. The mismatch in spatial scales between field inventory plots and larger pixels of current and planned remote sensing products for forest biomass mapping is of particular concern, as it has the potential to introduce errors, especially if forest biomass shows strong local spatial variation. Here, we used 30 large (8--50 ha) globally distributed permanent forest plots to quantify the spatial variability in aboveground biomass (AGB) at spatial grains ranging from 5 to 250 m (0.025--6.25 ha), and we evaluate the implications of this variability for calibrating remote sensing products using simulated remote sensing footprints. We found that the spatial sampling error in AGB is large for standard plot sizes, averaging 46.3{\%} for 0.1 ha subplots and 16.6{\%} for 1 ha subplots. Topographically heterogeneous sites showed positive spatial autocorrelation in AGB at scales of 100 m and above; at smaller scales, most study sites showed negative or nonexistent spatial autocorrelation in AGB. We further show that when field calibration plots are smaller than the remote sensing pixels, the high local spatial variability in AGB leads to a substantial {\textquotedbl}dilution{\textquotedbl} bias in calibration parameters, a bias that cannot be removed with current statistical methods. Overall, our results suggest that topography should be explicitly accounted for in future sampling strategies and that much care must be taken in designing calibration schemes if remote sensing of forest carbon is to achieve its promise.},
 author = {R{\'e}jou-M{\'e}chain, M. and Muller-Landau, H. C. and Detto, M. and Thomas, S. C. and {Le Toan}, T. and Saatchi, S. S. and Barreto-Silva, J. S. and Bourg, N. A. and Bunyavejchewin, S. and Butt, N. and Brockelman, W. Y. and Cao, M. and C{\'a}rdenas, D. and Chiang, J. M. and Chuyong, G. B. and Clay, K. and Condit, R. and Dattaraja, H. S. and Davies, S. J. and Duque, A. and Esufali, S. and Ewango, C. and Fernando, R. H. S. and Fletcher, C. D. and {Gunatilleke, I. A. U. N.} and Hao, Z. and Harms, K. E. and Hart, T. B. and H{\'e}rault, B. and Howe, R. W. and Hubbell, S. P. and Johnson, D. J. and Kenfack, D. and Larson, A. J. and Lin, L. and Lin, Y. and Lutz, J. A. and Makana, J. R. and Malhi, Y. and Marthews, T. R. and McEwan, R. W. and McMahon, S. M. and McShea, W. J. and Muscarella, R. and Nathalang, A. and Noor, N. S. M. and Nytch, C. J. and Oliveira, A. A. and Phillips, R. P. and Pongpattananurak, N. and Punchi-Manage, Ruwan and Salim, R. and Schurman, J. and Sukumar, R. and Suresh, H. S. and Suwanvecho, U. and Thomas, D. W. and Thompson, J. and Ur$\backslash$'{\i}arte, M. and Valencia, R. and Vicentini, A. and Wolf, A. T. and Yap, S. and Yuan, Z. and Zartman, C. E. and Zimmerman, J. K. and Chave, J.},
 year = {2014},
 title = {Local spatial structure of forest biomass and its consequences for remote sensing of carbon stocks},
 url = {http://dx.doi.org/10.5194/bgd-11-5711-2014},
 keywords = {ecol;phd},
 pages = {5711--5742},
 volume = {11},
 number = {4},
 journal = {Biogeosciences Discussions},
 doi = {10.5194/bgd-11-5711-2014},
 howpublished = {refereed}
}

Advances in forest carbon mapping have the potential to greatly reduce uncertainties in the global carbon budget and to facilitate effective emissions mitigation strategies such as REDD+. Though broad scale mapping is based primarily on remote sensing data, the accuracy of resulting forest carbon stock estimates depends critically on the quality of field measurements and calibration procedures. The mismatch in spatial scales between field inventory plots and larger pixels of current and planned remote sensing products for forest biomass mapping is of particular concern, as it has the potential to introduce errors, especially if forest biomass shows strong local spatial variation. Here, we used 30 large (8–50 ha) globally distributed permanent forest plots to quantify the spatial variability in aboveground biomass (AGB) at spatial grains ranging from 5 to 250 m (0.025–6.25 ha), and we evaluate the implications of this variability for calibrating remote sensing products using simulated remote sensing footprints. We found that the spatial sampling error in AGB is large for standard plot sizes, averaging 46.3% for 0.1 ha subplots and 16.6% for 1 ha subplots. Topographically heterogeneous sites showed positive spatial autocorrelation in AGB at scales of 100 m and above; at smaller scales, most study sites showed negative or nonexistent spatial autocorrelation in AGB. We further show that when field calibration plots are smaller than the remote sensing pixels, the high local spatial variability in AGB leads to a substantial \textquotedbldilution\textquotedbl bias in calibration parameters, a bias that cannot be removed with current statistical methods. Overall, our results suggest that topography should be explicitly accounted for in future sampling strategies and that much care must be taken in designing calibration schemes if remote sensing of forest carbon is to achieve its promise.

@article{Saefken2014Unifying,
 author = {S{\"a}fken, Benjamin and Kneib, Thomas and {van Waveren}, Clara-Sophie and Greven, Sonja},
 year = {2014},
 title = {A Unifying Approach to the Estimation of the Conditional Akaike Information in Generalized Linear Mixed Models},
 url = {http://dx.doi.org/10.1214/14-EJS881},
 keywords = {phd;stat},
 pages = {201--225},
 volume = {8},
 issn = {19357524},
 journal = {Electronic Journal of Statistics},
 doi = {10.1214/14-EJS881},
 howpublished = {refereed}
}

@article{Weigend2014Biodiversitat,
 author = {Weigend, S. and Jan{\ss}en-Tapken, U. and Erbe, Malena and Ober, U. and Weigend, A. and Preisinger, R. and Simianer, H.},
 year = {2014},
 title = {{B}iodiversit{\"a}t beim {Huhn} --- {Potenziale} f{\"u}r die {Praxis}},
 keywords = {phd},
 pages = {24--41},
 volume = {86},
 number = {1},
 journal = {Z{\"u}chtungskunde},
 howpublished = {refereed}
}

@article{Zhang2014Improving,
 abstract = {Utilizing the whole genomic variation of complex traits to predict the yet-to-be observed phenotypes or unobserved genetic values via whole genome prediction (WGP) and to infer the underlying genetic architecture via genome wide association study (GWAS) is an interesting and fast developing area in the context of human disease studies as well as in animal and plant breeding. Though thousands of significant loci for several species were detected via GWAS in the past decade, they were not used directly to improve WGP due to lack of proper models. Here, we propose a generalized way of building trait-specific genomic relationship matrices which can exploit GWAS results in WGP via a best linear unbiased prediction (BLUP) model for which we suggest the name BLUP|GA. Results from two illustrative examples show that using already existing GWAS results from public databases in BLUP|GA improved the accuracy of WGP for two out of the three model traits in a dairy cattle data set, and for nine out of the 11 traits in a rice diversity data set, compared to the reference methods GBLUP and BayesB. While BLUP|GA outperforms BayesB, its required computing time is comparable to GBLUP. Further simulation results suggest that accounting for publicly available GWAS results is potentially more useful for WGP utilizing smaller data sets and/or traits of low heritability, depending on the genetic architecture of the trait under consideration. To our knowledge, this is the first study incorporating public GWAS results formally into the standard GBLUP model and we think that the BLUP|GA approach deserves further investigations in animal breeding, plant breeding as well as human genetics.},
 author = {Zhang, Zhe and Ober, Ulrike and Erbe, Malena and Zhang, Hao and Gao, Ning and He, Jinlong and Li, Jiaqi and Simianer, Henner},
 year = {2014},
 title = {Improving the accuracy of whole genome prediction for complex traits using the results of genome wide association studies},
 url = {http://dx.doi.org/10.1371/journal.pone.0093017},
 keywords = {gen;phd},
 pages = {e93017},
 volume = {9},
 number = {3},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0093017},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/24663104},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3963961},
 howpublished = {refereed}
}

Utilizing the whole genomic variation of complex traits to predict the yet-to-be observed phenotypes or unobserved genetic values via whole genome prediction (WGP) and to infer the underlying genetic architecture via genome wide association study (GWAS) is an interesting and fast developing area in the context of human disease studies as well as in animal and plant breeding. Though thousands of significant loci for several species were detected via GWAS in the past decade, they were not used directly to improve WGP due to lack of proper models. Here, we propose a generalized way of building trait-specific genomic relationship matrices which can exploit GWAS results in WGP via a best linear unbiased prediction (BLUP) model for which we suggest the name BLUP|GA. Results from two illustrative examples show that using already existing GWAS results from public databases in BLUP|GA improved the accuracy of WGP for two out of the three model traits in a dairy cattle data set, and for nine out of the 11 traits in a rice diversity data set, compared to the reference methods GBLUP and BayesB. While BLUP|GA outperforms BayesB, its required computing time is comparable to GBLUP. Further simulation results suggest that accounting for publicly available GWAS results is potentially more useful for WGP utilizing smaller data sets and/or traits of low heritability, depending on the genetic architecture of the trait under consideration. To our knowledge, this is the first study incorporating public GWAS results formally into the standard GBLUP model and we think that the BLUP|GA approach deserves further investigations in animal breeding, plant breeding as well as human genetics.

@article{Beckschafer2013Enhanced,
 abstract = {The horizontal distribution of stems, stand density and the differentiation of tree dimensions are among the most important aspects of stand structure. An increasing complexity of stand structure is often linked to a higher number of species and to greater ecological stability. For quantification, the Structural Complexity Index (SCI) describes structural complexity by means of an area ratio of the surface that is generated by connecting the tree tops of neighbouring trees to form triangles to the surface that is covered by all triangles if projected on a flat plane. Here, we propose two ecologically relevant modifications of the SCI: The degree of mingling of tree attributes, quantified by a vector ruggedness measure, and a stem density term. We investigate how these two modifications influence index values. Data come from forest inventory field plots sampled along a disturbance gradient from heavily disturbed shrub land, through secondary regrowth to mature montane rainforest stands in Mengsong, Xishuangbanna,Yunnan,China. An application is described linking structural complexity, as described by the SCI and its modified versions, to changes in species composition of insect communities. The results of this study show that the Enhanced Structural Complexity Index (ESCI) can serve as a valuable tool for forest managers and ecologists for describing the structural complexity of forest stands and is particularly valuable for natural forests with a high degree of structural complexity.},
 author = {Becksch{\"a}fer, Philip and Mundhenk, Philip and Kleinn, Christoph and Ji, Yinqiu and Yu, Douglas W. and Harrison, Rhett D.},
 year = {2013},
 title = {Enhanced Structural Complexity Index: An Improved Index for Describing Forest Structural Complexity},
 url = {http://dx.doi.org/10.4236/ojf.2013.31005},
 keywords = {ecol;phd},
 pages = {23--29},
 volume = {3},
 number = {01},
 issn = {2163-0429},
 journal = {Open Journal of Forestry},
 doi = {10.4236/ojf.2013.31005},
 howpublished = {refereed}
}

The horizontal distribution of stems, stand density and the differentiation of tree dimensions are among the most important aspects of stand structure. An increasing complexity of stand structure is often linked to a higher number of species and to greater ecological stability. For quantification, the Structural Complexity Index (SCI) describes structural complexity by means of an area ratio of the surface that is generated by connecting the tree tops of neighbouring trees to form triangles to the surface that is covered by all triangles if projected on a flat plane. Here, we propose two ecologically relevant modifications of the SCI: The degree of mingling of tree attributes, quantified by a vector ruggedness measure, and a stem density term. We investigate how these two modifications influence index values. Data come from forest inventory field plots sampled along a disturbance gradient from heavily disturbed shrub land, through secondary regrowth to mature montane rainforest stands in Mengsong, Xishuangbanna,Yunnan,China. An application is described linking structural complexity, as described by the SCI and its modified versions, to changes in species composition of insect communities. The results of this study show that the Enhanced Structural Complexity Index (ESCI) can serve as a valuable tool for forest managers and ecologists for describing the structural complexity of forest stands and is particularly valuable for natural forests with a high degree of structural complexity.

@article{Beduschi2013Friends,
 abstract = {The importance of ontogeny and the degree of abiotic stress in determining the interplay between facilitation and competition is well known. However, their joint effect on the outcome of plant interactions remains poorly understood, especially when a continuous gradient of abiotic stress is considered. Our objective was to evaluate the frequency of association of individuals of Clusia criuva with typical coastal dune species across a gradient of water stress and how this association affects the growth of juveniles and sub-adults. The study was performed in a coastal dune region in South Brazil, where the sandy soil promotes severe water stress. One-year growth of 293 individuals and their distance to the closest humid slacks were measured. This distance is a good surrogate for water stress, since slacks represent proximity to groundwater. The proportion of associated individuals increased with abiotic stress in both ontogenetic stages, but was always greater for juveniles. This suggests that association is progressively more important to guarantee survival as abiotic stress increases. Nonetheless, the benefit of neighbors to growth decreased with abiotic stress, and associated plants grew less than isolated ones in harsher environments. This was mainly true for juveniles, since the height growth of sub-adults was not affected by association or abiotic stress. In our study, facilitation became more intense with environmental severity, increasing survival, although competition also became more influent, reducing growth particularly for younger plants. This demonstrates that ontogenetic stage and abiotic stress must be considered simultaneously in order to better understand interactions among plants.},
 author = {Beduschi, Tatiane and Castellani, T{\^a}nia T.},
 year = {2013},
 title = {Friends or Foes? {I}nterplay of facilitation and competition depends on the interaction between abiotic stress and ontogenetic stage},
 url = {http://dx.doi.org/10.1007/s11258-013-0269-8},
 keywords = {ecol;phd},
 pages = {1485--1492},
 volume = {214},
 number = {12},
 journal = {Plant Ecology},
 doi = {10.1007/s11258-013-0269-8},
 howpublished = {refereed}
}

The importance of ontogeny and the degree of abiotic stress in determining the interplay between facilitation and competition is well known. However, their joint effect on the outcome of plant interactions remains poorly understood, especially when a continuous gradient of abiotic stress is considered. Our objective was to evaluate the frequency of association of individuals of Clusia criuva with typical coastal dune species across a gradient of water stress and how this association affects the growth of juveniles and sub-adults. The study was performed in a coastal dune region in South Brazil, where the sandy soil promotes severe water stress. One-year growth of 293 individuals and their distance to the closest humid slacks were measured. This distance is a good surrogate for water stress, since slacks represent proximity to groundwater. The proportion of associated individuals increased with abiotic stress in both ontogenetic stages, but was always greater for juveniles. This suggests that association is progressively more important to guarantee survival as abiotic stress increases. Nonetheless, the benefit of neighbors to growth decreased with abiotic stress, and associated plants grew less than isolated ones in harsher environments. This was mainly true for juveniles, since the height growth of sub-adults was not affected by association or abiotic stress. In our study, facilitation became more intense with environmental severity, increasing survival, although competition also became more influent, reducing growth particularly for younger plants. This demonstrates that ontogenetic stage and abiotic stress must be considered simultaneously in order to better understand interactions among plants.

@article{Chisholm2013Scaledependent,
 author = {Chisholm, Ryan A. and Muller-Landau, Helene C. and {Abdul Rahman}, Kassim and Bebber, Daniel P. and Bin, Yue and Bohlman, Stephanie A. and Bourg, Norman A. and Brinks, Joshua and Bunyavejchewin, Sarayudh and Butt, Nathalie and Cao, Honglin and Cao, Min and C{\'a}rdenas, Dairon and Chang, Li-Wan and Chiang, Jyh-Min and Chuyong, George and Condit, Richard and Dattaraja, Handanakere S. and Davies, Stuart and Duque, Alvaro and Fletcher, Christine and Gunatilleke, Nimal and Gunatilleke, Savitri and Hao, Zhanqing and Harrison, Rhett D. and Howe, Robert and Hsieh, Chang-Fu and Hubbell, Stephen P. and Itoh, Akira and Kenfack, David and Kiratiprayoon, Somboon and Larson, Andrew J. and Lian, Juyu and Lin, Dunmei and Liu, Haifeng and Lutz, James A. and Ma, Keping and Malhi, Yadvinder and McMahon, Sean and McShea, William and Meegaskumbura, Madhava and Mohd and Morecroft, Michael D. and Nytch, Christopher J. and Oliveira, Alexandre and Parker, Geoffrey G. and Pulla, Sandeep and Punchi-Manage, Ruwan and Romero-Saltos, Hugo and Sang, Weiguo and Schurman, Jon and Su, Sheng-Hsin and Sukumar, Raman and Sun, I-Fang and Suresh, Hebbalalu S. and Tan, Sylvester and Thomas, Duncan and Thomas, Sean and Thompson, Jill and Valencia, Renato and Wolf, Amy and Yap, Sandra and Ye, Wanhui and Yuan, Zuoqiang and Zimmerman, Jess K.},
 year = {2013},
 title = {Scale-dependent relationships between tree species richness and ecosystem function in forests},
 url = {http://dx.doi.org/10.1111/1365-2745.12132},
 keywords = {ecol;phd},
 pages = {1214--1224},
 volume = {101},
 number = {5},
 issn = {00220477},
 journal = {Journal of Ecology},
 doi = {10.1111/1365-2745.12132},
 howpublished = {refereed}
}

@article{Erbe2013function,
 abstract = {Prediction of genomic breeding values is of major practical relevance in dairy cattle breeding. Deterministic equations have been suggested to predict the accuracy of genomic breeding values in a given design which are based on training set size, reliability of phenotypes, and the number of independent chromosome segments ([Formula: see text]). The aim of our study was to find a general deterministic equation for the average accuracy of genomic breeding values that also accounts for marker density and can be fitted empirically. Two data sets of 5'698 Holstein Friesian bulls genotyped with 50 K SNPs and 1'332 Brown Swiss bulls genotyped with 50 K SNPs and imputed to $\sim$600 K SNPs were available. Different k-fold (k = 2-10, 15, 20) cross-validation scenarios (50 replicates, random assignment) were performed using a genomic BLUP approach. A maximum likelihood approach was used to estimate the parameters of different prediction equations. The highest likelihood was obtained when using a modified form of the deterministic equation of Daetwyler et al. (2010), augmented by a weighting factor (w) based on the assumption that the maximum achievable accuracy is [Formula: see text]. The proportion of genetic variance captured by the complete SNP sets ([Formula: see text]) was 0.76 to 0.82 for Holstein Friesian and 0.72 to 0.75 for Brown Swiss. When modifying the number of SNPs, w was found to be proportional to the log of the marker density up to a limit which is population and trait specific and was found to be reached with $\sim$20'000 SNPs in the Brown Swiss population studied.},
 author = {Erbe, Malena and Gredler, Birgit and Seefried, Franz Reinhold and Bapst, Beat and Simianer, Henner},
 year = {2013},
 title = {A function accounting for training set size and marker density to model the average accuracy of genomic prediction},
 keywords = {PI},
 pages = {e81046},
 volume = {8},
 number = {12},
 journal = {PLOS ONE},
 doi = {10.1371/journal.pone.0081046},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/24339895},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3855218},
 howpublished = {refereed}
}

Prediction of genomic breeding values is of major practical relevance in dairy cattle breeding. Deterministic equations have been suggested to predict the accuracy of genomic breeding values in a given design which are based on training set size, reliability of phenotypes, and the number of independent chromosome segments ([Formula: see text]). The aim of our study was to find a general deterministic equation for the average accuracy of genomic breeding values that also accounts for marker density and can be fitted empirically. Two data sets of 5'698 Holstein Friesian bulls genotyped with 50 K SNPs and 1'332 Brown Swiss bulls genotyped with 50 K SNPs and imputed to $∼$600 K SNPs were available. Different k-fold (k = 2-10, 15, 20) cross-validation scenarios (50 replicates, random assignment) were performed using a genomic BLUP approach. A maximum likelihood approach was used to estimate the parameters of different prediction equations. The highest likelihood was obtained when using a modified form of the deterministic equation of Daetwyler et al. (2010), augmented by a weighting factor (w) based on the assumption that the maximum achievable accuracy is [Formula: see text]. The proportion of genetic variance captured by the complete SNP sets ([Formula: see text]) was 0.76 to 0.82 for Holstein Friesian and 0.72 to 0.75 for Brown Swiss. When modifying the number of SNPs, w was found to be proportional to the log of the marker density up to a limit which is population and trait specific and was found to be reached with $∼$20'000 SNPs in the Brown Swiss population studied.

@article{Freytag2014Comparison,
 author = {Freytag, Saskia and Bickeb{\"o}ller, Heike},
 year = {2013},
 title = {Comparison of Three Summary Statistics for Ranking Genes in Genome-Wide Association Studies},
 url = {http://dx.doi.org/10.1002/sim.6063},
 keywords = {gen;phd},
 journal = {Statistics in Medicine},
 doi = {10.1002/sim.6063},
 howpublished = {refereed}
}

@article{Freytag2014Network,
 abstract = {Biological pathways provide rich information and biological context on the genetic causes of complex diseases. The logistic kernel machine test integrates prior knowledge on pathways in order to analyze data from genome-wide association studies (GWAS). In this study, the kernel converts the genomic information of 2 individuals into a quantitative value reflecting their genetic similarity. With the selection of the kernel, one implicitly chooses a genetic effect model. Like many other pathway methods, none of the available kernels accounts for the topological structure of the pathway or gene-gene interaction types. However, evidence indicates that connectivity and neighborhood of genes are crucial in the context of GWAS, because genes associated with a disease often interact. Thus, we propose a novel kernel that incorporates the topology of pathways and information on interactions. Using simulation studies, we demonstrate that the proposed method maintains the type I error correctly and can be more effective in the identification of pathways associated with a disease than non-network-based methods. We apply our approach to genome-wide association case-control data on lung cancer and rheumatoid arthritis. We identify some promising new pathways associated with these diseases, which may improve our current understanding of the genetic mechanisms.},
 author = {Freytag, Saskia and Manitz, Juliane and Schlather, Martin and Kneib, Thomas and {[...]} and Bickeb{\"o}ller, Heike},
 year = {2013},
 title = {A network-based kernel machine test for the identification of risk pathways in genome-wide association studies},
 url = {http://dx.doi.org/DOI:10.1159/000357567},
 keywords = {gen;phd},
 pages = {64--75},
 volume = {76},
 number = {2},
 journal = {Human Heredity},
 doi = {10.1159/000357567},
 file = {http://www.ncbi.nlm.nih.gov/pubmed/24434848},
 file = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4026009},
 howpublished = {refereed}
}

Biological pathways provide rich information and biological context on the genetic causes of complex diseases. The logistic kernel machine test integrates prior knowledge on pathways in order to analyze data from genome-wide association studies (GWAS). In this study, the kernel converts the genomic information of 2 individuals into a quantitative value reflecting their genetic similarity. With the selection of the kernel, one implicitly chooses a genetic effect model. Like many other pathway methods, none of the available kernels accounts for the topological structure of the pathway or gene-gene interaction types. However, evidence indicates that connectivity and neighborhood of genes are crucial in the context of GWAS, because genes associated with a disease often interact. Thus, we propose a novel kernel that incorporates the topology of pathways and information on interactions. Using simulation studies, we demonstrate that the proposed method maintains the type I error correctly and can be more effective in the identification of pathways associated with a disease than non-network-based methods. We apply our approach to genome-wide association case-control data on lung cancer and rheumatoid arthritis. We identify some promising new pathways associated with these diseases, which may improve our current understanding of the genetic mechanisms.

@article{Kormann2013Sabine,
 author = {Kormann, U. and Sturmberger, B.},
 year = {2013},
 title = {A {Sabine's Gull} \textit{Xema sabini} in {Montenegro} -- first record in the north-eastern mediterranean and a short review of the species' status in the region},
 keywords = {phd},
 pages = {79--82},
 volume = {34},
 journal = {Acrocephalus},
 howpublished = {refereed}
}

@article{Manitz2013Bayesian,
 author = {Manitz, Juliane and H{\"o}hle, Michael},
 year = {2013},
 title = {Bayesian outbreak detection algorithm for monitoring reported cases of campylobacteriosis in {Germany}},
 url = {http://dx.doi.org/10.1002/bimj.201200141},
 keywords = {phd;stat},
 pages = {509--526},
 volume = {55},
 number = {4},
 issn = {03233847},
 journal = {Biometrical Journal},
 doi = {10.1002/bimj.201200141},
 howpublished = {refereed}
}

@article{Martini2013Interaction,
 abstract = {The decoupled sites representation (DSR) for one type of ligand allows to regard complex overall titration curves as sum of classical Henderson-Hasselbalch (HH) titration curves. In this work we transfer this theoretical approach to molecules with different types of interacting ligands (e.g. protons and electrons), prove the existence of decoupled systems for n

1 and one binding sites for two different ligands, and point out some difficulties and limits of this transfer. A major difference to the DSR for one type of ligand is the loss of uniqueness of the decoupled system. However, all decoupled systems share a unique set of microstate probabilities and each decoupled system corresponds to a certain permutation of these microstate probabilities. Moreover, we show that the titration curve of a certain binding site in the original system can be regarded as linear combination of the titration curves of the individual sites of the decoupled system if the weights of the linear combination are substituted by functions in the activity of the second ligand. In the underlying model with only pairwise interaction, an important observation of our theoretical investigation is the following: Even though the binding sites of ligand L

1 may not interact directly, they can show secondary interaction due to the interaction with the second type of ligand. This means, if the activity of the second ligand is fixed and we regard the 1-dimensional titration curve of an individual binding site for ligand L

1 depending on its activity, we may observe a strong deviation from the classical HH shape in spite of non-interacting sites for ligand L

1.},
 author = {Martini, Johannes W. R. and Schlather, Martin and Ullmann, G. Matthias},
 year = {2013},
 title = {On the interaction of two different types of ligands binding to the same molecule part I: basics and the transfer of the decoupled sites representation to systems with n and one binding sites},
 url = {http://dx.doi.org/10.1007/s10910-012-0107-6},
 keywords = {phd},
 pages = {672--695},
 volume = {51},
 number = {2},
 issn = {0259-9791},
 journal = {Journal of Mathematical Chemistry},
 doi = {10.1007/s10910-012-0107-6},
 howpublished = {refereed}
}

The decoupled sites representation (DSR) for one type of ligand allows to regard complex overall titration curves as sum of classical Henderson-Hasselbalch (HH) titration curves. In this work we transfer this theoretical approach to molecules with different types of interacting ligands (e.g. protons and electrons), prove the existence of decoupled systems for n 1 and one binding sites for two different ligands, and point out some difficulties and limits of this transfer. A major difference to the DSR for one type of ligand is the loss of uniqueness of the decoupled system. However, all decoupled systems share a unique set of microstate probabilities and each decoupled system corresponds to a certain permutation of these microstate probabilities. Moreover, we show that the titration curve of a certain binding site in the original system can be regarded as linear combination of the titration curves of the individual sites of the decoupled system if the weights of the linear combination are substituted by functions in the activity of the second ligand. In the underlying model with only pairwise interaction, an important observation of our theoretical investigation is the following: Even though the binding sites of ligand L 1 may not interact directly, they can show secondary interaction due to the interaction with the second type of ligand. This means, if the activity of the second ligand is fixed and we regard the 1-dimensional titration curve of an individual binding site for ligand L 1 depending on its activity, we may observe a strong deviation from the classical HH shape in spite of non-interacting sites for ligand L 1.

@article{Martini2013InteractionII,
 abstract = {In the first part of this work we formulated the decoupled sites representation for two different types of ligands and highlighted special properties of the case of n binding sites for ligand L

1 and one binding site for ligand L

2. Moreover, for this case, we identified the microstate constants as unique components all decoupled molecules share. In the second part on hand, we investigate the cases with (n, 2) and (n, 3) binding sites. As it is difficult to solve the system of equations occurring when a molecule with more than one binding site for both ligands shall be decoupled, we present applicable calculation methods which exploit the special structure of the system of equations. Moreover, we investigate which unique properties all decoupled molecules share and show that for two different decoupled molecules with the same binding polynomial, not all microstate constants of a certain macrostate are permutations of the microstate constants of the other molecule.},
 author = {Martini, Johannes W. R. and Schlather, Martin and Ullmann, G. Matthias},
 year = {2013},
 title = {On the interaction of different types of ligands binding to the same molecule Part II: systems with n to 2 and n to 3 binding sites},
 url = {http://dx.doi.org/10.1007/s10910-012-0108-5},
 keywords = {phd},
 pages = {696--714},
 volume = {51},
 number = {2},
 issn = {0259-9791},
 journal = {Journal of Mathematical Chemistry},
 doi = {10.1007/s10910-012-0108-5},
 howpublished = {refereed}
}

In the first part of this work we formulated the decoupled sites representation for two different types of ligands and highlighted special properties of the case of n binding sites for ligand L 1 and one binding site for ligand L 2. Moreover, for this case, we identified the microstate constants as unique components all decoupled molecules share. In the second part on hand, we investigate the cases with (n, 2) and (n, 3) binding sites. As it is difficult to solve the system of equations occurring when a molecule with more than one binding site for both ligands shall be decoupled, we present applicable calculation methods which exploit the special structure of the system of equations. Moreover, we investigate which unique properties all decoupled molecules share and show that for two different decoupled molecules with the same binding polynomial, not all microstate constants of a certain macrostate are permutations of the microstate constants of the other molecule.

@article{Martini2013Meaning,
 author = {Martini, Johannes W. R. and Ullmann, G. M. and Schlather, M.},
 year = {2013},
 title = {The Meaning of the Decoupled Sites Representation in Terms of Statistical Mechanics and Stochastics},
 url = {http://match.pmf.kg.ac.rs/electronic\textunderscore versions/Match70/n3/match70n3\textunderscore 829-850.pdf},
 keywords = {phd},
 pages = {829--850},
 volume = {70},
 number = {3},
 journal = {MATCH Communications in Mathematical and in Computer Chemistry},
 howpublished = {refereed}
}

@article{PunchiManage2013Effects,
 abstract = {* One of the primary goals in community ecology is to determine the relative importance of processes and mechanisms that control biodiversity. Here, we examined habitat-driven species assemblages and species distribution patterns as well as their temporal variations for three life stages of two censuses of a 25-ha mixed dipterocarp forest at Sinharaja (Sri Lanka). * Our general objective was to find out whether the species assemblages and associated habitat types changed with life stage, spatial scale and species attributes. We also analyse whether the habitat types were related to certain indicator species. Habitat types were determined with multivariate regression tree analyses driven by topographic variables. * We found species assemblages associated with five distinct habitat types that appeared consistently for all life stages of the two censuses. These habitats were related to ridge-valley gradients and a pronounced contrast in south-west versus north-east aspect. Habitat-driven structuring was weak at the recruit stage but strong in the juvenile and adult stages. The species assemblage variance explained by topographic variables for different life stages ranged between 10{\%} for recruits and 23{\%} for juveniles. * The species assemblages determined for different spatial scales (10, 20, 50~m) showed similar habitat partitioning, but the variance explained by the topographic variables increased in all life stages with spatial scale. This could be due to the homogenizing effect of topographic variables at the larger scales and unaccounted environmental variation at the smaller scales. The number of indicator species identified in the two censuses was higher in the juvenile stage than in the adult stage, and nearly all indicator species in the adult stage were also indicator species in the juvenile stage. * Synthesis.~Our study showed that approximately 75{\%} of the variance in local species composition is unexplained. This may be due to spatially structured processes such as dispersal limitation, unaccounted biotic and abiotic environmental variables, and stochastic effects, but only 25{\%} were due to topographic habitat association. Although the pronounced ridge-valley gradient and contrast of south-west versus north-east aspect created consistent habitats, our results suggest that local species assemblages at Sinharaja forest are jointly shaped by neutral and niche processes.},
 author = {Punchi-Manage, Ruwan and Getzin, Stephan and Wiegand, Thorsten and Kanagaraj, Rajapandian and {Savitri Gunatilleke}, C. V. and {Nimal Gunatilleke, I. A. U.} and Wiegand, Kerstin and Huth, Andreas},
 year = {2013},
 title = {Effects of topography on structuring local species assemblages in a {Sri Lankan} mixed dipterocarp forest},
 url = {http://dx.doi.org/10.1111/1365-2745.12017},
 keywords = {ecol;phd},
 pages = {149--160},
 volume = {101},
 number = {1},
 issn = {00220477},
 journal = {Journal of Ecology},
 doi = {10.1111/1365-2745.12017},
 howpublished = {refereed}
}

* One of the primary goals in community ecology is to determine the relative importance of processes and mechanisms that control biodiversity. Here, we examined habitat-driven species assemblages and species distribution patterns as well as their temporal variations for three life stages of two censuses of a 25-ha mixed dipterocarp forest at Sinharaja (Sri Lanka). * Our general objective was to find out whether the species assemblages and associated habitat types changed with life stage, spatial scale and species attributes. We also analyse whether the habitat types were related to certain indicator species. Habitat types were determined with multivariate regression tree analyses driven by topographic variables. * We found species assemblages associated with five distinct habitat types that appeared consistently for all life stages of the two censuses. These habitats were related to ridge-valley gradients and a pronounced contrast in south-west versus north-east aspect. Habitat-driven structuring was weak at the recruit stage but strong in the juvenile and adult stages. The species assemblage variance explained by topographic variables for different life stages ranged between 10% for recruits and 23% for juveniles. * The species assemblages determined for different spatial scales (10, 20, 50 m) showed similar habitat partitioning, but the variance explained by the topographic variables increased in all life stages with spatial scale. This could be due to the homogenizing effect of topographic variables at the larger scales and unaccounted environmental variation at the smaller scales. The number of indicator species identified in the two censuses was higher in the juvenile stage than in the adult stage, and nearly all indicator species in the adult stage were also indicator species in the juvenile stage. * Synthesis. Our study showed that approximately 75% of the variance in local species composition is unexplained. This may be due to spatially structured processes such as dispersal limitation, unaccounted biotic and abiotic environmental variables, and stochastic effects, but only 25% were due to topographic habitat association. Although the pronounced ridge-valley gradient and contrast of south-west versus north-east aspect created consistent habitats, our results suggest that local species assemblages at Sinharaja forest are jointly shaped by neutral and niche processes.

@article{Sobotka2013Confidence,
 abstract = {In regression scenarios there is a growing demand for information on the conditional distribution of the response beyond the mean. In this scenario quantile regression is an established method of tail analysis. It is well understood in terms of asymptotic properties and estimation quality. Another way to look at the tail of a distribution is via expectiles. They provide a valuable alternative since they come with a combination of preferable attributes. The easy weighted least squares estimation of expectiles and the quadratic penalties often used in flexible regression models are natural partners. Also, in a similar way as quantiles can be seen as a generalisation of median regression, expectiles offer a generalisation of mean regression. In addition to regression estimates, confidence intervals are essential for interpretational purposes and to assess the variability of the estimate, but there is a lack of knowledge regarding the asymptotic properties of a semiparametric expectile regression estimate. Therefore confidence intervals for expectiles based on an asymptotic normal distribution are introduced. Their properties are investigated by a simulation study and compared to a boostrap-based gold standard method. Finally the introduced confidence intervals help to evaluate a geoadditive expectile regression model on childhood malnutrition data from India.},
 author = {Sobotka, Fabian and Kauermann, G{\"o}ran and {Schulze Waltrup}, Linda and Kneib, Thomas},
 year = {2013},
 title = {On confidence intervals for semiparametric expectile regression},
 url = {http://dx.doi.org/10.1007/s11222-011-9297-1},
 keywords = {postdoc;stat},
 pages = {135--148},
 volume = {23},
 number = {2},
 issn = {0960-3174},
 journal = {Statistics and Computing},
 doi = {10.1007/s11222-011-9297-1},
 howpublished = {refereed}
}

In regression scenarios there is a growing demand for information on the conditional distribution of the response beyond the mean. In this scenario quantile regression is an established method of tail analysis. It is well understood in terms of asymptotic properties and estimation quality. Another way to look at the tail of a distribution is via expectiles. They provide a valuable alternative since they come with a combination of preferable attributes. The easy weighted least squares estimation of expectiles and the quadratic penalties often used in flexible regression models are natural partners. Also, in a similar way as quantiles can be seen as a generalisation of median regression, expectiles offer a generalisation of mean regression. In addition to regression estimates, confidence intervals are essential for interpretational purposes and to assess the variability of the estimate, but there is a lack of knowledge regarding the asymptotic properties of a semiparametric expectile regression estimate. Therefore confidence intervals for expectiles based on an asymptotic normal distribution are introduced. Their properties are investigated by a simulation study and compared to a boostrap-based gold standard method. Finally the introduced confidence intervals help to evaluate a geoadditive expectile regression model on childhood malnutrition data from India.

@article{Sobotka2013Estimating,
 author = {Sobotka, Fabian and Radice, Rosalba and Marra, Giampiero and Kneib, Thomas},
 year = {2013},
 title = {Estimating the relationship between women's education and fertility in {Botswana} by using an instrumental variable approach to semiparametric expectile regression},
 url = {http://dx.doi.org/10.1111/j.1467-9876.2012.01050.x},
 keywords = {postdoc;stat},
 pages = {25--45},
 volume = {62},
 number = {1},
 issn = {00359254},
 journal = {Journal of the Royal Statistical Society: Series C (Applied Statistics)},
 doi = {10.1111/j.1467-9876.2012.01050.x},
 howpublished = {refereed}
}

@article{Freytag2012Novel,
 abstract = {The logistic kernel machine test (LKMT) is a testing procedure tailored towards high-dimensional genetic data. Its use in pathway analyses of case-control genome-wide association studies results from its computational efficiency and flexibility in incorporating additional information via the kernel. The kernel can be any positive definite function; unfortunately, its form strongly influences the test's power and bias. Most authors have recommended the use of a simple linear kernel. We demonstrate via a simulation that the probability of rejecting the null hypothesis of no association just by chance increases with the number of SNPs or genes in the pathway when applying a simple linear kernel.

We propose a novel kernel that includes an appropriate standardization in order to protect against any inflation of false positive results. Moreover, our novel kernel contains information on gene membership of SNPs in the pathway.

When applying the novel kernel to data from the North American Rheumatoid Arthritis Consortium, we find that even this basic genomic structure can improve the ability of the LKMT to identify meaningful associations. We also demonstrate that the standardization effectively eliminates problems of size bias.

We recommend the use of our standardized kernel and urge caution when using non-adjusted kernels in the LKMT to conduct pathway analyses.

Copyright {\copyright} 2013 S. Karger AG, Basel.},
 author = {Freytag, Saskia and Bickeb{\"o}ller, Heike and Amos, Christopher I. and Kneib, Thomas and Schlather, Martin},
 year = {2012},
 title = {A novel kernel for correcting size bias in the logistic kernel machine test with an application to rheumatoid arthritis},
 url = {http://dx.doi.org/10.1159/000347188},
 keywords = {gen;phd},
 pages = {97--108},
 volume = {74},
 number = {2},
 journal = {Human Heredity},
 doi = {10.1159/000347188},
 howpublished = {refereed}
}

The logistic kernel machine test (LKMT) is a testing procedure tailored towards high-dimensional genetic data. Its use in pathway analyses of case-control genome-wide association studies results from its computational efficiency and flexibility in incorporating additional information via the kernel. The kernel can be any positive definite function; unfortunately, its form strongly influences the test's power and bias. Most authors have recommended the use of a simple linear kernel. We demonstrate via a simulation that the probability of rejecting the null hypothesis of no association just by chance increases with the number of SNPs or genes in the pathway when applying a simple linear kernel. We propose a novel kernel that includes an appropriate standardization in order to protect against any inflation of false positive results. Moreover, our novel kernel contains information on gene membership of SNPs in the pathway. When applying the novel kernel to data from the North American Rheumatoid Arthritis Consortium, we find that even this basic genomic structure can improve the ability of the LKMT to identify meaningful associations. We also demonstrate that the standardization effectively eliminates problems of size bias. We recommend the use of our standardized kernel and urge caution when using non-adjusted kernels in the LKMT to conduct pathway analyses. Copyright © 2013 S. Karger AG, Basel.

@article{Herrmann2012Effects,
 abstract = {Habitat isolation is expected to reduce population densities of animals via reduced immigration. However, altered trophic interactions in isolated habitats may modify these effects, especially since the strength of isolation effects is expected to increase with trophic rank. Here, we studied effects of habitat isolation on a food-web module consisting of herbivorous beetles, predatory spiders, spider-preying wasps and arthropod-feeding birds. We compare two systems that were studied in subsequent years: a study on 29 mature apple orchards that varied in the degree of isolation from forest, and a study on 20 groups of newly planted cherry trees that showed similar variation in their degree of habitat isolation. No birds were observed on the young fruit trees. Wasps and spiders showed the expected lower abundances in isolated habitats. On mature trees, birds were present and showed lower abundances in isolated habitats. Wasps were reduced to a similar degree by habitat isolation as on the young trees. Surprisingly, spider densities on the mature trees were higher in isolated than in connected habitat. This contrasting response of spiders to habitat isolation is likely to be due to release from bird predation in isolated mature orchards. In both study systems, beetles showed no significant effect of habitat isolation. Our results confirm that the sensitivity to fragmentation increases with trophic rank, and suggest that trophic interactions should receive more attention in fragmentation studies.},
 author = {Herrmann, J. and Kormann, U. and Sch{\"u}epp, C. and Stocker, Y. and Herzog, F. and Entling, M.},
 year = {2012},
 title = {Effects of habitat isolation and predation pressure on an arboreal food-web},
 url = {http://dx.doi.org/10.1556/comec.13.2012.1.10},
 keywords = {ecol;phd},
 pages = {82--87},
 volume = {13},
 number = {1},
 journal = {Community Ecology},
 doi = {10.1556/comec.13.2012.1.10},
 howpublished = {refereed}
}

Habitat isolation is expected to reduce population densities of animals via reduced immigration. However, altered trophic interactions in isolated habitats may modify these effects, especially since the strength of isolation effects is expected to increase with trophic rank. Here, we studied effects of habitat isolation on a food-web module consisting of herbivorous beetles, predatory spiders, spider-preying wasps and arthropod-feeding birds. We compare two systems that were studied in subsequent years: a study on 29 mature apple orchards that varied in the degree of isolation from forest, and a study on 20 groups of newly planted cherry trees that showed similar variation in their degree of habitat isolation. No birds were observed on the young fruit trees. Wasps and spiders showed the expected lower abundances in isolated habitats. On mature trees, birds were present and showed lower abundances in isolated habitats. Wasps were reduced to a similar degree by habitat isolation as on the young trees. Surprisingly, spider densities on the mature trees were higher in isolated than in connected habitat. This contrasting response of spiders to habitat isolation is likely to be due to release from bird predation in isolated mature orchards. In both study systems, beetles showed no significant effect of habitat isolation. Our results confirm that the sensitivity to fragmentation increases with trophic rank, and suggest that trophic interactions should receive more attention in fragmentation studies.

@article{Kormann2012Parsimonybased,
 abstract = {Connectivity and dispersal are key components determining the persistence of fragmented populations. However, the assessment of dispersal rates and landscape factors affecting functional connectivity remains problematic for species living in small populations, even more so in stress-sensitive, rare or elusive species. In this study, we estimated current dispersal patterns in a regional population of the endangered capercaillie (Tetrao urogallus) using a novel, parsimony-based application of pedigree analysis. In spring 2003 and 2008, we systematically collected non-invasive samples of this forest bird species in five local populations in the north-eastern Pre-Alps of Switzerland. Based on 11 nuclear microsatellites, we identified 86 unique genotypes over the two sampling periods and determined levels of genetic structure. We investigated the relative impact of topography and land use on functional population connectivity by comparing inter-individual genetic distance with GIS-derived least-cost path models representing different dispersal scenarios. Results show that local populations are connected by dispersers, suggesting occasional contemporary gene flow. Mountain ridges were found to hamper capercaillie dispersal, while, at odds with our hypothesis, neither a broad valley nor land use type notably affected gene flow. These results suggest that our study region represents a patchy metapopulation. Our study illustrates how pedigree analysis and landscape genetic methods can significantly contribute to the understanding of dispersal and connectivity in rare and elusive species. {\^a}–º Noninvasive genetic assessment of dispersal and functional connectivity in capercaillie (Tetrao urogallus). {\^a}–º Least-cost path analysis suggest mountains, but not land use types to hamper dispersal. {\^a}–º New parsimony-based pedigree analysis shows frequent dispersal among local populations. {\^a}–º Study region represent a patchy metapopulation. {\^a}–º New approach seems suitable for other rare and elusive, species of conservation concern.},
 author = {Kormann, Urs and Gugerli, Felix and Ray, Nicolas and Excoffier, Laurent and Bollmann, Kurt},
 year = {2012},
 title = {Parsimony-based pedigree analysis and individual-based landscape genetics suggest topography to restrict dispersal and connectivity in the endangered capercaillie},
 url = {http://dx.doi.org/10.1016/j.biocon.2012.04.011},
 keywords = {ecol;phd},
 pages = {241--252},
 volume = {152},
 issn = {00063207},
 journal = {Biological Conservation},
 doi = {10.1016/j.biocon.2012.04.011},
 howpublished = {refereed}
}

Connectivity and dispersal are key components determining the persistence of fragmented populations. However, the assessment of dispersal rates and landscape factors affecting functional connectivity remains problematic for species living in small populations, even more so in stress-sensitive, rare or elusive species. In this study, we estimated current dispersal patterns in a regional population of the endangered capercaillie (Tetrao urogallus) using a novel, parsimony-based application of pedigree analysis. In spring 2003 and 2008, we systematically collected non-invasive samples of this forest bird species in five local populations in the north-eastern Pre-Alps of Switzerland. Based on 11 nuclear microsatellites, we identified 86 unique genotypes over the two sampling periods and determined levels of genetic structure. We investigated the relative impact of topography and land use on functional population connectivity by comparing inter-individual genetic distance with GIS-derived least-cost path models representing different dispersal scenarios. Results show that local populations are connected by dispersers, suggesting occasional contemporary gene flow. Mountain ridges were found to hamper capercaillie dispersal, while, at odds with our hypothesis, neither a broad valley nor land use type notably affected gene flow. These results suggest that our study region represents a patchy metapopulation. Our study illustrates how pedigree analysis and landscape genetic methods can significantly contribute to the understanding of dispersal and connectivity in rare and elusive species. ► Noninvasive genetic assessment of dispersal and functional connectivity in capercaillie (Tetrao urogallus). ► Least-cost path analysis suggest mountains, but not land use types to hamper dispersal. ► New parsimony-based pedigree analysis shows frequent dispersal among local populations. ► Study region represent a patchy metapopulation. ► New approach seems suitable for other rare and elusive, species of conservation concern.

@article{Martini2012Mathematical,
 abstract = {The decoupled sites representation (DSR) is a theoretical instrument which allows to regard complex pH titration curves of biomolecules with several interacting proton binding sites as composition of isolated, non-interacting sites, each with a standard Henderson--Hasselbalch titration curve. In this work, we present the mathematical framework in which the DSR is embedded and give mathematical proofs for several statements in the periphery of the DSR. These proofs also identify exceptions. To apply the DSR to any molecule, it is necessary to extend the set of binding energies from R to a stripe within C. An important observation in this context is that even positive interaction energies (repulsion) between the binding sites will not guarantee real binding energies in the decoupled system, at least if the molecule has more than four proton binding sites. Moreover, we show that for a given overall titration curve it is not only possible to find a corresponding system with an interaction energy of zero but with any arbitrary fix interaction energy. This result also effects practical work as it shows that for any given titration curve, there is an infinite number of corresponding hypothetical molecules. Furthermore, this implies that---using a common definition of cooperative binding on the level of interaction energies---a meaningful measure of cooperativity between the binding sites cannot be defined solely on the basis of the overall titration. Consequently, all measures of cooperativity based on the overall binding curve do not measure the type of cooperativity commonly defined on the basis of interaction energies. Understanding the DSR mathematically provides the basis of transferring the DSR to biomolecules with different types of interacting ligands, such as protons and electrons, which play an important role within electron transport chains like in photosynthesis.},
 author = {Martini, Johannes W. R. and Ullmann, G. Matthias},
 year = {2012},
 title = {A mathematical view on the decoupled sites representation},
 url = {http://dx.doi.org/10.1007/s00285-012-0517-x},
 keywords = {phd},
 pages = {477--503},
 volume = {66},
 issn = {0303-6812},
 journal = {Journal of Mathematical Biology},
 doi = {10.1007/s00285-012-0517-x},
 howpublished = {refereed}
}

The decoupled sites representation (DSR) is a theoretical instrument which allows to regard complex pH titration curves of biomolecules with several interacting proton binding sites as composition of isolated, non-interacting sites, each with a standard Henderson–Hasselbalch titration curve. In this work, we present the mathematical framework in which the DSR is embedded and give mathematical proofs for several statements in the periphery of the DSR. These proofs also identify exceptions. To apply the DSR to any molecule, it is necessary to extend the set of binding energies from R to a stripe within C. An important observation in this context is that even positive interaction energies (repulsion) between the binding sites will not guarantee real binding energies in the decoupled system, at least if the molecule has more than four proton binding sites. Moreover, we show that for a given overall titration curve it is not only possible to find a corresponding system with an interaction energy of zero but with any arbitrary fix interaction energy. This result also effects practical work as it shows that for any given titration curve, there is an infinite number of corresponding hypothetical molecules. Furthermore, this implies that—using a common definition of cooperative binding on the level of interaction energies—a meaningful measure of cooperativity between the binding sites cannot be defined solely on the basis of the overall titration. Consequently, all measures of cooperativity based on the overall binding curve do not measure the type of cooperativity commonly defined on the basis of interaction energies. Understanding the DSR mathematically provides the basis of transferring the DSR to biomolecules with different types of interacting ligands, such as protons and electrons, which play an important role within electron transport chains like in photosynthesis.

@article{Ober2012Using,
 abstract = {Predicting organismal phenotypes from genotype data is important for plant and animal breeding, medicine, and evolutionary biology. Genomic-based phenotype prediction has been applied for single-nucleotide polymorphism (SNP) genotyping platforms, but not using complete genome sequences. Here, we report genomic prediction for starvation stress resistance and startle response in Drosophila melanogaster, using $\sim$2.5 million SNPs determined by sequencing the Drosophila Genetic Reference Panel population of inbred lines. We constructed a genomic relationship matrix from the SNP data and used it in a genomic best linear unbiased prediction (GBLUP) model. We assessed predictive ability as the correlation between predicted genetic values and observed phenotypes by cross-validation, and found a predictive ability of 0.239$\pm$0.008 (0.230$\pm$0.012) for starvation resistance (startle response). The predictive ability of BayesB, a Bayesian method with internal SNP selection, was not greater than GBLUP. Selection of the 5{\%} SNPs with either the highest absolute effect or variance explained did not improve predictive ability. Predictive ability decreased only when fewer than 150,000 SNPs were used to construct the genomic relationship matrix. We hypothesize that predictive power in this population stems from the SNP--based modeling of the subtle relationship structure caused by long-range linkage disequilibrium and not from population structure or SNPs in linkage disequilibrium with causal variants. We discuss the implications of these results for genomic prediction in other organisms. The ability to accurately predict values of complex phenotypes from genotype data will revolutionize plant and animal breeding, personalized medicine, and evolutionary biology. To date, genomic prediction has utilized high-density single-nucleotide polymorphism (SNP) genotyping arrays, but the availability of sequence data opens new frontiers for genomic prediction methods. This article is the first application of genomic phenotype prediction using whole-genome sequence data in a substantial sample of a higher eukaryote. We use $\sim$2.5 million SNPs with minor allele frequency greater than 2.5{\%} derived from genomic sequences of the  '' Drosophila Genetic Reference Panel'' to predict phenotypes for two traits, starvation resistance and startle-induced locomotor behavior. We systematically address prediction within versus across sexes, genomic best linear unbiased prediction (GBLUP) versus a Bayesian approach, and the effect of SNP density. We find that (i) genomic prediction can be efficiently implemented using sequence data via GBLUP, (ii) there is little gain in predictive ability if the number of SNPs is increased above 150,000, and (iii) neither implicit nor explicit marker selection substantially improves the predictive ability. Although the findings must be seen against the background of small sample sizes, the results illustrate both the potential of the approach and the challenges ahead.},
 author = {Ober, Ulrike and {[...]} and Schlather, Martin and Mackay, Trudy F. C. and Simianer, Henner},
 year = {2012},
 title = {Using whole-genome sequence data to predict quantitative trait phenotypes in \textit{Drosophila melanogaster}},
 url = {http://dx.doi.org/10.1371/journal.pgen.1002685},
 keywords = {gen;phd},
 pages = {e1002685},
 volume = {8},
 number = {5},
 journal = {PLoS Genetics},
 doi = {10.1371/journal.pgen.1002685},
 howpublished = {refereed}
}

Predicting organismal phenotypes from genotype data is important for plant and animal breeding, medicine, and evolutionary biology. Genomic-based phenotype prediction has been applied for single-nucleotide polymorphism (SNP) genotyping platforms, but not using complete genome sequences. Here, we report genomic prediction for starvation stress resistance and startle response in Drosophila melanogaster, using $∼$2.5 million SNPs determined by sequencing the Drosophila Genetic Reference Panel population of inbred lines. We constructed a genomic relationship matrix from the SNP data and used it in a genomic best linear unbiased prediction (GBLUP) model. We assessed predictive ability as the correlation between predicted genetic values and observed phenotypes by cross-validation, and found a predictive ability of 0.239$±$0.008 (0.230$±$0.012) for starvation resistance (startle response). The predictive ability of BayesB, a Bayesian method with internal SNP selection, was not greater than GBLUP. Selection of the 5% SNPs with either the highest absolute effect or variance explained did not improve predictive ability. Predictive ability decreased only when fewer than 150,000 SNPs were used to construct the genomic relationship matrix. We hypothesize that predictive power in this population stems from the SNP–based modeling of the subtle relationship structure caused by long-range linkage disequilibrium and not from population structure or SNPs in linkage disequilibrium with causal variants. We discuss the implications of these results for genomic prediction in other organisms. The ability to accurately predict values of complex phenotypes from genotype data will revolutionize plant and animal breeding, personalized medicine, and evolutionary biology. To date, genomic prediction has utilized high-density single-nucleotide polymorphism (SNP) genotyping arrays, but the availability of sequence data opens new frontiers for genomic prediction methods. This article is the first application of genomic phenotype prediction using whole-genome sequence data in a substantial sample of a higher eukaryote. We use $∼$2.5 million SNPs with minor allele frequency greater than 2.5% derived from genomic sequences of the '' Drosophila Genetic Reference Panel'' to predict phenotypes for two traits, starvation resistance and startle-induced locomotor behavior. We systematically address prediction within versus across sexes, genomic best linear unbiased prediction (GBLUP) versus a Bayesian approach, and the effect of SNP density. We find that (i) genomic prediction can be efficiently implemented using sequence data via GBLUP, (ii) there is little gain in predictive ability if the number of SNPs is increased above 150,000, and (iii) neither implicit nor explicit marker selection substantially improves the predictive ability. Although the findings must be seen against the background of small sample sizes, the results illustrate both the potential of the approach and the challenges ahead.

@article{Ober2011Predicting,
 abstract = {Genomic data provide a valuable source of information for modeling covariance structures, allowing a more accurate prediction of total genetic values (GVs). We apply the kriging concept, originally developed in the geostatistical context for predictions in the low-dimensional space, to the high-dimensional space spanned by genomic single nucleotide polymorphism (SNP) vectors and study its properties in different gene-action scenarios. Two different kriging methods [{\textquotedbl}universal kriging{\textquotedbl} (UK) and {\textquotedbl}simple kriging{\textquotedbl} (SK)] are presented. As a novelty, we suggest use of the family of Mat{\'e}rn covariance functions to model the covariance structure of SNP vectors. A genomic best linear unbiased prediction (GBLUP) is applied as a reference method. The three approaches are compared in a whole-genome simulation study considering additive, additive-dominance, and epistatic gene-action models. Predictive performance is measured in terms of correlation between true and predicted GVs and average true GVs of the individuals ranked best by prediction. We show that UK outperforms GBLUP in the presence of dominance and epistatic effects. In a limiting case, it is shown that the genomic covariance structure proposed by VanRaden (2008) can be considered as a covariance function with corresponding quadratic variogram. We also prove theoretically that if a specific linear relationship exists between covariance matrices for two linear mixed models, the GVs resulting from BLUP are linked by a scaling factor. Finally, the relation of kriging to other models is discussed and further options for modeling the covariance structure, which might be more appropriate in the genomic context, are suggested.},
 author = {Ober, Ulrike and Erbe, Malena and Long, Nanye and Porcu, Emilio and Schlather, Martin and Simianer, Henner},
 year = {2011},
 title = {Predicting genetic values: A kernel-based best linear unbiased prediction with genomic data},
 url = {http://dx.doi.org/10.1534/genetics.111.128694},
 keywords = {gen;phd},
 pages = {695--708},
 volume = {188},
 number = {3},
 issn = {1943-2631},
 journal = {Genetics},
 doi = {10.1534/genetics.111.128694},
 howpublished = {refereed}
}

Genomic data provide a valuable source of information for modeling covariance structures, allowing a more accurate prediction of total genetic values (GVs). We apply the kriging concept, originally developed in the geostatistical context for predictions in the low-dimensional space, to the high-dimensional space spanned by genomic single nucleotide polymorphism (SNP) vectors and study its properties in different gene-action scenarios. Two different kriging methods [\textquotedbluniversal kriging\textquotedbl (UK) and \textquotedblsimple kriging\textquotedbl (SK)] are presented. As a novelty, we suggest use of the family of Matérn covariance functions to model the covariance structure of SNP vectors. A genomic best linear unbiased prediction (GBLUP) is applied as a reference method. The three approaches are compared in a whole-genome simulation study considering additive, additive-dominance, and epistatic gene-action models. Predictive performance is measured in terms of correlation between true and predicted GVs and average true GVs of the individuals ranked best by prediction. We show that UK outperforms GBLUP in the presence of dominance and epistatic effects. In a limiting case, it is shown that the genomic covariance structure proposed by VanRaden (2008) can be considered as a covariance function with corresponding quadratic variogram. We also prove theoretically that if a specific linear relationship exists between covariance matrices for two linear mixed models, the GVs resulting from BLUP are linked by a scaling factor. Finally, the relation of kriging to other models is discussed and further options for modeling the covariance structure, which might be more appropriate in the genomic context, are suggested.

@article{Muhlner2010Structural,
 author = {M{\"u}hlner, Susanne and Kormann, Urs and Schmidt-Entling, Martin and Herzog, Felix and Bailey, Debra},
 year = {2010},
 title = {Structural Versus Functional Habitat Connectivity Measures to Explain Bird Diversity in Fragmented Orchards},
 url = {http://dx.doi.org/10.2478/v10285-012-0023-2},
 keywords = {ecol;phd},
 pages = {52--64},
 volume = {3},
 number = {1},
 issn = {1805-4196},
 journal = {Journal of Landscape Ecology},
 doi = {10.2478/v10285-012-0023-2},
 howpublished = {refereed}
}