Concordance between male- and female-specific GWAS results helps define underlying genetic architecture of complex traits

A better understanding of genetic architecture will help enhance precision medicine and clinical care. Towards this end, we investigate sex-stratified analyses for several traits in the Hybrid Mouse Diversity Panel (HMDP) and UK Biobank to assess trait polygenicity and identify contributing loci. By comparing allelic effect directions in males and females, we hypothesize that non-associated loci should show random effect directions across sexes. Instead, we observe strong concordance in effect direction, even among alleles lacking nominal statistical significance. Our findings suggest hundreds of loci influence each mouse trait and thousands affect each human trait, including traits with no significant loci under conventional approaches. We also detect patterns consistent with spurious widespread epistasis. These results highlight the value of sex-stratified analyses in uncovering novel loci, suggest a method for identifying biologically relevant associations beyond statistical thresholds, and caution that pervasive main effects may produce misleading epistatic signals.