A sex-specific genome-wide association study of blood lipid levels in All of Us

Figures

• 

  Figure 1.. Flowchart of cohort selection The figure outlines the sequential filtering of participants with whole-genome sequencing data, lipid measurements, and demographic information. After removing individuals with incomplete or invalid lipid values and those without complete paired lipid measurements, a final cohort of 124,920 participants (75,127 female; 49,793 male) was retained for genome-wide association analysis (GWAS) (srWGS: short-read whole genome sequencing data; HDL-C: high density lipoprotein cholesterol; LDL-C: low density lipoprotein cholesterol; TG: triglyceride; TC: total cholesterol; mg/dL: milligrams per deciliter; Others: participants whose sex-at-birth field contained “I prefer not to answer,” “pmi: skip,” “sex at birth none of these,” or “intersex”).

• 

  Figure 2.. Age and sex differential distribution of HDL-C, LDL-C, TG, and Total Cholesterol Density plots show HDL-C and LDL-C distributions across age groups for females and males. For each age stratum, sample sizes and Cohen’s d are reported where Cohen’s d quantifies the standardized mean difference between sexes using the following equation (positive d means higher mean in females).

• 

  Figure 3.. Manhattan plots for pooled GWAS Manhattan plots display pooled genome-wide association results for HDL-C (Figure 3A), LDL-C (Figure 3B), TG (Figure 3C), and TC (Figure 3D) in the pooled cohort. Each point represents a genetic variant plotted by chromosomal position and −log10(P). The horizontal dashed line denotes the genome-wide significance threshold after Bonferroni correction.

• 

  Figure 4.. Regional association plots showing sex-specific associations. Regional association plots illustrate sex-specific associations. Variants are shown with −log10(P) values plotted against genomic position, and are colored according to linkage disequilibrium (r2) with the lead variant. Gene models and transcript orientations are displayed below each plot to contextualize the association signals using the locus zoom package.

• 

  Figure 5.. rs7528419 at chr1:109274570:A:G shows male-specific significance for HDL-C (CELSR2). Boxplots show lipid profile levels stratified by genotype and sex for variants exhibiting sex-specific associations.Values outside of 1.5 inter-quartile range from quartiles were not shown. Significance threshold are noted in asterisks using the following convention: ****: Tuckey adjusted p-value < 0.0001 ***: Tuckey adjusted p-value < 0.001 **: Tuckey adjusted p-value < 0.01 *: Tuckey adjusted p-value < 0.05