Fig. 1
From: Improving accuracy of genomic prediction by genetic architecture based priors in a Bayesian model
Distribution of p-values and locus-specific π of three traits in dairy cattle population across the genome. Rows in the figure correspond to distributions of features of milk fat percentage (FP), milk yield (MY), and somatic cell score (SCS), respectively. Four columns correspond to distributions of ω, density of ω, distribution of locus-specific π, and density of locus-specific π, respectively; where, ω = − log 10(p − values). The p-values are derived from ANOVA for all single markers. Logarithmic transformation of the p-values is performed for data visualization convenience and latter utilization. The locus-specific π is derived from the p-values of the ANOVA via formula (4). Since π is the proportion of non-effective markers, 1-π is taken as the probability of each marker to be effective. For milk yield and milk fat percentage, the clusters on chromosome 14 is the genomic segment where located the DGAT1 gene. For somatic cell score, no cluster is observed due to the lack of major genes. Distributions of the locus-specific π are consistent with our prior knowledge about the genetic architectures of these traits. These plots are drawn on the R software platform (http://www.r-project.org/)