A genome-wide meta-analysis identifies a sex-specific genetic effect for platelet aggregation in response to agonists

Figures

• 

  Figure 1:. GWAS meta-analysis Manhattan plots testing the 1 degree of freedom SNP × Sex interaction for phenotypes EPI LOW 3 (top) and EPI LOW 4 (bottom). The dashed horizontal line is located at p = 5 × 10–8, representing the standard GWAS cut-off for significance. SNPs passing this significance threshold are highlighted with a red background and listed in Table 1.

• 

  Figure 2:. Violin plots for measured EPI LOW 3 percent aggregation (top row, a - c) and adjusted inverse normalized transformed EPI LOW 3 percent aggregation (bottom row, d - f) in rs116725046 reference homozygotes (non-carriers) and heterozygotes (carriers) female and male (red and blue, respectively). Shown are GeneSTAR participants of European ancestry (left, a and d), FHS participants (middle, b and e) and GeneSTAR participants of African ancestry (right, c and f). The observed number of participants in each group are shown at the top axis.

• 

  Figure 3:. FUMA circos plots for chromosome 10 with peak SNPs rs116725046 (EPI LOW 3, left) and rs141008381 (EPI LOW 4, right). The outer layer is a Manhattan plot showing −log10 p-values for SNPs with nominal significance (p < 0.05). SNPs with r2 > 0.8 to one of the significant loci identified are colored in red. Other SNPs are colored in gray. The outer circle (chartreuse color) shows the chromosome genomic coordinates in Mb. The risk loci are highlighted in blue. Genes mapped by either Hi-C or eQTLs are shown on the inner circle. Several genes were mapped by Hi-C and are colored orange, with the respective chromatin interactions shown as orange links. Only LINC00702 was mapped by both Hi-C and eQTLs and is colored in red. No genes were mapped by eQTLs alone.

• 

  Figure 4:. The z-statistics from sex-stratified LINC00702 eQTLs for SNPs rs116725046 and rs141008381. For each of 35 tissues, linear models were used to test an association between genotype and expression in males and females, respectively, adjusting for known sample and donor characteristics and surrogate variables that capture hidden technical or biological factors of expression variability (see (Oliva et al., 2020) for details). The z-statistic was defined as the slope estimate divided by its estimated standard error. Among females, no p-value lower than 0.01 were observed, corresponding to a z-statistic of 2.58 in absolute value for two-sided tests and indicated by the light blue region. However, 12 and 11 statistics among males yielded p-values lower than 0.01, respectively, with all slopes being negative, showing lower expression of LINC00702 for carriers of the respective variants.