Data Availability
Genotype data were processed using PLINK v2.00: https://www.cog-genomics.org/plink/1.9/; MAMBA R package: https://github.com/dan11mcguire/mamba; metafor R package v3.4.0: https://cran.r-project.org/package=metafor; Rcpp R package v1.0.9: https://cran.r-project.org/package=Rcpp; TMB R package v1.9.1: https://cran.r-project.org/package=TMB; figures were generated using ggplot2 R package v3.3.6: https://cran.r-project.org/package=ggplot2 and figures with break axis were generated using ggbreak package v0.1.1: https://cran.r-project.org/package=ggbreak; Manhattan plot was generated using qqman R package v0.1.8: https://cran.r-project.org/package=qqman; Venn diagram was generated using ggvenn R package v0.1.9: https://cran.r-project.org/package=ggvenn; upset plot was generated using UpSetR R package v1.4.0; GWAS Catalog: https://www.ebi.ac.uk/gwas; Open Targets: https://www.opentargets.org. Our method is implemented in the MAJAR R package, freely available at https://github.com/JustinaXie/MAJAR.