Identifying novel genetic loci associated with polycystic ovary syndrome based on its shared genetic architecture with type 2 diabetes

Figures

• 

  FIGURE 1. Flow diagram of the overall study design. GWAS summary statistics on PCOS and T2D with or without adjustment of BMI were collected from publicly available databases and open literature reports. We quantified the polygenic overlap between PCOS and T2D. Then, we visualized the genetic pleiotropy enrichment between both traits and identified the genetic loci associated with PCOS or/and T2D. Finally, four bioinformatics approaches were applied to explore the potential function of the identified genetic loci.

• 

  FIGURE 2. Venn diagrams of shared causal variants (gray). (A) T2D unadjusted by BMI (red) and PCOS (blue); (B) T2D adjusted by BMI (red) and PCOS (blue). The estimated numbers of causal variants are shown in thousands, explaining 90% of SNP heritability in each phenotype, followed by the standard error. The Dice coefficients, an overall measure on a 0–100% scale of the polygenic overlap, are 44.1% (A) and 32.1% (B), respectively. The genetic correlations ( rg ) estimated by the MiXeR model are 0.41 (A) and 0.21 (B), respectively.

• 

  FIGURE 3. Genetic pleiotropy enrichment of PCOS conditional on T2D (without adjustment for BMI) and vice versa. Stratified conditional Q–Q plot of nominal versus empirical and negative log10-transformed p-values in the primary phenotype as a functional significance of association with the secondary phenotype at the level of p < 1(All SNPs, blue), p < 0.1 (orange), p < 0.01 (green), and p < 0.001 (purple) corresponding to −log10P>0, −log10P>1, −log10P>2,and −log10P>3 , respectively. (A) PCOS conditioned on T2D; (B) T2D conditioned on PCOS. Dotted lines demonstrate the null hypothesis. The summary statistics of T2D without adjustment of BMI were selected for plotting those conditional Q–Q plots.

• 

  FIGURE 4. Manhattan plot showing 11 genetic loci associated with PCOS or/and T2D. Manhattan plots showing the negative log10-transformed condFDR values (A) and negative log10-transformed conjFDR values (B) for each SNP on the y axis and chromosomal positions along the x axis. The most significant SNP lead in each LD block is enlarged and encircled in black, whereas the small points represent other SNPs. Labels correspond to lead SNPs previously reported for published loci (black) and for newly identified loci (red).

• 

  FIGURE 5. Distribution of the functional annotation for candidate SNPs. (A) Distribution of the functional consequences of independent SNPs. (B) Distribution of RegulomeDB scores, with a lower score indicating a higher likelihood of having a regulatory function. (C) Minimum chromatin state across 127 tissue and cell types for SNPs in a shared genomic region, with lower states indicating higher accessibility and states 1–7 referring to open chromatin states.