Fig. 1: C-GWAS identified 188 genetic loci with study-wide significant face association.

a Localization of 44 landmarks on the face. b Eight facial regions and their corresponding landmarks. c QQ plots for C-GWAS combined p-values (one-sided), MinGWAS adjusted p-values (one-sided), and associated p-values of the most inflated single-trait GWAS (two-sided). Grey dots represent results from the GWAS for the facial trait L21_L23, which had the highest inflation among all single-trait GWASs. LDSC intercepts and \({\lambda }_{{{\rm{G}}}{{\rm{C}}}}\) for the three types of results are indicated. The solid line indicates the expected distribution of p-values under the null hypothesis. d Venn diagram illustrates the overlap between the 188 loci identified by C-GWAS and the 440 loci reported by previous GWASs obtained by our literature survey. e Miami plots display the results of C-GWAS combined p-values (one-sided) and MinGWAS adjusted p-values (one-sided). Solid and dashed lines represent the study-wide significance (\(5\times {10}^{-8}\)) and suggestive significance (\(1\times {10}^{-5}\)) thresholds, respectively. Loci with the study-wide significance are highlighted using three types of marks, including purple dots for overlapping loci between C-GWAS and MinGWAS, green crosses for independent loci from C-GWAS or MinGWAS, and orange triangles for novel loci identified by C-GWAS that were not reported in previous GWASs. The 3D template facial image in this figure is adapted from White et al.⁸² published under an Open Access license (CC BY 4.0), see http://creativecommons.org/licenses/by/4.0/.