Skip to main content
Fig. 1 | BMC Genomics

Fig. 1

From: A Bayesian model for detection of high-order interactions among genetic variants in genome-wide association studies

Fig. 1

Bayesian scheme in model relationship between traits of target population and genetic variations. a Traits are presented as Y, e.g., oil or protein content of soybeans, and genetic variations are presented as X, e.g., SNPs. X 1 to X 4 are the related binary genetic variations (green is 0 and yellow is 1), and they are inferred as group 1, which is associate with phenotype (Y). b Based on values of quantitative trait, Y can be divided into 4 clusters: cyan, black, blue and red, each corresponding to one circle. Within each cluster, Y follows a Gaussian distribution. The four Gaussian distributions can have different means and variances. In this case, X 1 to X 4, four of R genetic variations X = {X 1,…,X R} can be divided into 4 independent clusters of combination configurations (0011, 1010, 1100 and 1110), and they have a clear pattern associated with Y. Hence, the combination of X 1 , X 2, X 3 and X 4 can be treated as one genetic variation interaction. In contrast, if the genotype clusters overlap with each other significantly in the phenotype space, there is no evidence for such a genetic variation interaction

Back to article page