Bayesian Multi-Phenotype Genome-Wide Association Methods for Experimental Designs of Arbitrary Complexity

Genome-wide association studies (GWAS) in rice serve as an important starting point for breeding programs.  To maximize their potential, it is necessary to evaluate phenotypes that are relevant in target field environments.  Thus, complex experiments with a large environmental influence are often required.  However, these experiments produce noisy data that may obscure the contribution of genetic variation.  Most modern statistical methods for GWAS have been developed to map human disease loci, where each genetically-distinct individual is evaluated only once and the confounding effects of population structure are not strong.  These methods are not adequate for most applications in rice and other crop species.  We have developed a general Bayesian approach that accounts for noise in the data, including low heritability, unbalanced designs, missingness, outlier observations, and systematic experimental bias. We also implemented steps to correct for deep population structure (e.g., differentiation between O. s. indica and japonica), and analyze multiple phenotypes at once.  We illustrate the advantages of our methods on simulated data and on a set of agronomic phenotypes measured on a diversity panel of rice accessions as well as a set of maize breeding populations from CIMMYT.