2016/03/29 by Deniz Akdemir, Akdemir, Deniz, Jean‐Luc Jannink +1
Agricultural and Biological Sciences · Biochemistry, Genetics and Molecular Biology · #Applications (stat.AP) #FOS: Biological sciences #FOS: Computer and information sciences #Genetic Mapping and Diversity in Plants and Animals #Genetic and phenotypic traits in livestock #Genetics and Plant Breeding #Machine Learning (stat.ML) #Quantitative Methods (q-bio.QM)
paper · pdf · doi:10.48550/arxiv.1603.08813
openalex publication_date 2016/03/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In statistical genetics an important task involves building predictive models\nfor the genotype-phenotype relationships and thus attribute a proportion of the\ntotal phenotypic variance to the variation in genotypes. Numerous models have\nbeen proposed to incorporate additive genetic effects into models for\nprediction or association. However, there is a scarcity of models that can\nadequately account for gene by gene or other forms of genetical interactions.\nIn addition, there is an increased interest in using marker annotations in\ngenome-wide prediction and association. In this paper, we discuss an hybrid\nmodeling methodology which combines the parametric mixed modeling approach and\nthe non-parametric rule ensembles. This approach gives us a flexible class of\nmodels that can be used to capture additive, locally epistatic genetic effects,\ngene x background interactions and allows us to incorporate one or more\nannotations into the genomic selection or association models. We use benchmark\ndata sets covering a range of organisms and traits in addition to simulated\ndata sets to illustrate the strengths of this approach. The improvement of\nmodel accuracies and association results suggest that a part of the "missing\nheritability" in complex traits can be captured by modeling local epistasis.\n