Sex-stratified logistic mixed models: 42 binary traits

Sex-stratified GWAS can help shed light on sexual differences in genetic architecture. In Bernabeu et al (2021) we fit sex-stratified logistic mixed models (using REGENIE) across a total of 42 binary traits for which we’d found evidence of sexual differences for at least one genetic variant. Here, the results of said models, as well as the comparison statistics between male and female genetic effects, are included. Each file contains the results for a single binary trait, as stated in the file name, using its corresponding UK Biobank trait code. Trait descriptions, including their respective UK Biobank codes, are stated in the “trait_description_binary.tsv” file. For each trait (each .gz file), GWAS summary statistics obtained for over 9 million genetic variants across the autosomal genome and circa 450K individuals, as well as the results of the t-test comparing genetic effect estimates between the sexes, are included. ; This DataShare deposit includes GWAS datasets for 42 binary traits. Each file represents a trait, as stated in the file title. A table is also included which matches each trait code with a description. Fields include those returned by REGENIE for each sex: chromosome, position, ID, reference allele (allele 0), alternative allele (allele 1), frequency of the alternative allele, sample size, SNP imputation score and the test performed (additive). These are followed by the estimated effect sizes (first order approximation), standard errors (first order approximation), chi-square test statistic and -log10 p value. Finally, sex comparison statistics are included, including the t-statistic [STAT-Comp_REGENIE], as calculated on effects transformed on linear scale using FOA transformation, p-value [PVAL-Comp_REGENIE] and FDR corrected p-value [PVAL-Comp_REGENIE_FDR].