Fig. 4: Manhattan plots and quantile-quantile (QQ) plots of GWAS results for serum ferritin. | Nature Communications

Fig. 4: Manhattan plots and quantile-quantile (QQ) plots of GWAS results for serum ferritin.

From: SPAGRM: effectively controlling for sample relatedness in large-scale genome-wide association studies of longitudinal traits

Fig. 4: Manhattan plots and quantile-quantile (QQ) plots of GWAS results for serum ferritin.

a Mirror Manhattan plots of GWAS results of SPAGRM, SPAGRM(CCT), NormGRM, and TrajGWAS. b QQ plots in which genetic variants are grouped based on minor allele frequency (MAF): common variants with MAF in (0.05, 0.5), low-frequency variants with MAF in (0.01, 0.05), rare variants with MAF in (0.002, 0.01), and ultra-rare variants with MAF in (2e-4, 0.002). c, Scatterplots comparing SPAGRM and SPAGRM(CCT) with TrajGWAS on common variants (i.e., MAF > 0.05). Across all subplots, upper and lower panels are results for testing for mean profile (\({\beta }_{g}=0\)) and WS variability (\({\tau }_{g}=0\)), respectively. The red dashed line represents p value of \(5\times {10}^{-8}\), and the grey dashed line represents the breakpoint \({10}^{-15}\) for the different scales of the y-axis. For serum ferritin, 66,729 subjects were used for NormGRM, SPAGRM, and SPAGRM(CCT) analysis, of which 55,683 (83.4%) were used for TrajGWAS. An average of 2.26 values were measured for one subject. P value is calculated using two-sided score tests.

Back to article page