Estimating the Relative Contribution of Putatively Deleterious and Neutral SNPs to Agronomic Phenotypes

Targeted identification and purging of segregating deleterious alleles has been proposed as a novel approach to plant breeding. Recent advances in DNA resequencing technology and sequence constraint based approaches to predict the functional impact of a mutation now allow for identification of putatively deleterious SNPs on a genome-wide scale. Previous surveys of bioinformatically-identified deleterious SNPs show that individual crop genomes may carry hundreds of putatively deleterious SNPs. However, the contributions of these SNPs to phenotypic variation have not been empirically evaluated. We use exome resequencing and SNP genotyping of three cycles of a spring six-row barley breeding population to compare the phenotypic contributions of putatively deleterious SNPs to those of neutral SNPs. In this population, grain yield remained stable while a proxy for disease severity decreased. We find that putatively deleterious SNPs explain more phenotypic variance for yield, but less for disease severity on average than other functional classes of variants.