Fig. 4From: Estimating the effect size of a hidden causal factor between SNPs and a continuous trait: a mediation model approachHistograms of the regression estimated effects. (A) Estimated SNP effects on BMI (a∗). (B) Estimated SNP effects on BMI with outliers removed. (C) Estimated direct effects between the SNPs and WHR (c∗). (D) Estimated direct effects between the SNPs and WHR with outliers removedBack to article page