Extended Data Fig. 4: Forest plots for loss-of-function and predicted-deleterious missense variants in FAM189A2, CTNNA3 and GATA4.

This forest plot shows the contributions of each cohort and meta-analysis results for gene-based analysis of LOF and deleterious missense variants in FAM189A2, CTNNA3, and GATA4; center measures show odds ratios, while the error bars represent 95% confidence intervals. The odds ratio and 95% confidence interval of each study were estimated from Firth’s logistic regression, and the meta-analysis odds ratios and confidence intervals were estimated from an inverse-variance weighted fixed effect approach. The P-values of the meta-analysis were computed from a score-based meta-analysis approach; P-values are two-sided and unadjusted for multiple-testing. For CTNNA3, all datasets contributed (52,416 cases and 267,772 controls); for the other genes, TOPMed-CCDG, CCDG-WES, and UKBB contributed (51,019 cases and 253,267 controls); the cMAC shows the number of alternative alleles observed in the meta-analysis. cMAC, cumulative minor allele count; Combined, meta-analysis results.