Genetic Modifiers of Cardiac Remodeling Severity in IVS4+919G>A-Associated Fabry Cardiomyopathy

Figures

• 

  Figure 1. Genome-wide association analysis of dichotomized plasma globotriaosylsphingosine (lyso-Gb3) status and descriptive genotype–phenotype comparisons. (A) Manhattan plot of the GWAS for dichotomized plasma lyso-Gb3 status (≥2.6 versus <2.6 nM). Genomic position is shown on the x-axis and −log10(p) on the y-axis. Selected top-ranked variants are labeled. The horizontal red line marks the post hoc exploratory cutoff (p ≤ 1 × 10−5), not genome-wide significance. The arrows indicate the corresponding labeled top-ranked variants. (B) Q–Q plot of observed and expected −log10(p) values. (C–H) Genotype-specific distributions of continuous plasma lyso-Gb3 for selected top-ranked SNPs: (C) rs7671863 (p = 0.0037); (D) rs11181117 (p = 0.0158); (E) rs7689578 (p = 0.0178); (F) rs150794871 (p = 0.0264); (G) rs3191122 (p = 0.0311); and (H) rs3740423 (p = 0.0344). Horizontal lines denote group means. Panels C–H are descriptive continuous-trait comparisons based on the Mann–Whitney U test. Their p values differ from the Cochran–Armitage trend test results for the dichotomized GWAS phenotype in Table 2.

• 

  Figure 2. Genome-wide association analysis of left ventricular hypertrophy status and genotype-specific left ventricular mass index distributions. (A) Manhattan plot of the GWAS for LVH status. The horizontal line marks the original study-specific significance threshold of p = 1 × 10−7. rs1435166 and rs2572260 met this threshold and the conventional p < 5 × 10−8 threshold. The arrows indicate the two top-ranked variants, rs1538664 and rs2749713. (B) Q–Q plot of observed and expected −log10(p) values. The original analysis reported λGC = 1.00, which cannot exclude residual population stratification. (C–H) Genotype-specific LVMI distributions: (C) rs1435166 genotypes (p = 0.0006); (D) rs1435166 under a dominant model, C-allele carriers versus T/T (p < 0.0001); (E) rs2749713 genotypes (p = 0.0012); (F) rs2749713 under a dominant model (p = 0.0003); (G) rs1538664 genotypes (p = 0.0003); and (H) rs1538664 under a dominant model, T-allele carriers versus C/C (p < 0.0001). Horizontal lines denote mean LVMI. Panels C–H are descriptive follow-up comparisons. Three-group comparisons used the Kruskal–Wallis test with Dunn’s post hoc test, and two-group comparisons used the Mann–Whitney U test. These…

• 

  Figure 3. Exploratory genome-wide association analysis of septal hypertrophy status. (A) Manhattan plot of the GWAS for septal hypertrophy status. rs12612117 and rs17778833 met the exploratory p ≤ 1 × 10−6 threshold used in the original analysis but not the conventional genome-wide threshold of p < 5 × 10−8. The arrows indicate the two top-ranked variants, rs12612117 and rs17778833. (B) Q–Q plot of observed and expected p-value distributions. (C,D) Genotype-specific IVSd distributions for rs12612117: (C) by genotype (p = 0.0153) and (D) under a dominant model, A/A versus G/A + G/G (p = 0.0110). Horizontal lines denote mean IVSd. No genotype-level plot is shown for rs17778833.