Fig. 5: Per-allele squared effect size of height as a function of MAF.
From: Efficient variance components analysis across millions of genomes
We applied RHE-mc to N = 291,273 unrelated white British individuals and M = 7,774,235 imputed SNPs. SNPs were partitioned into 144 bins based on LD score (4 bins based on quartiles of the LD score with i denoting the ith quartile) and MAF (36 MAF bins) (see “Methods” section). Per-allele effect size squared for bin k is defined as \(\frac{{h}_{k}^{2}}{{M}_{k}* 2{f}_{k}* (1\,-\,{f}_{k})}\), where \({h}_{k}^{2}\) is the heritability attributed to bin k, Mk is the number of SNPs in bin k, and fk is the average MAF in bin k. Each point represents the estimated allelic effect size. Bars mark ±2 standard errors centered on the estimated allelic effect size. See Supplementary Fig. 9 for results on all 22 traits.
