Fig. 3: lo-siRF finds epistatic interactions between genetic risk loci for left ventricular hypertrophy.

a, Circos plot showing the genetic risk loci identified by lo-siRF (green, circle 2) and regions after clumping FUMA-extracted SNVs in LD (r² > 0.6) with lo-siRF-prioritized SNVs (black, circle 3). Circle 1 shows the top 300 epistatic SNV-SNV pairs with the highest frequency of occurrence in lo-siRF (green), SNV-gene linkages (FDR < 0.5) based on GTEx⁴¹ V8 cis-eQTL information from heart and skeletal muscle tissues (purple), and 3D chromatin interactions³⁷ based on Hi-C data of left ventricular tissue obtained from GSE87112. Circle 5 and 6 show bar plots of the occurrence frequency and number of partner SNVs in epistasis (normalized by the maximum value of the corresponding locus) identified by lo-siRF, respectively. Circle 7 shows the ChromHMM³⁸ core-15 chromatin state for left ventricle (LV), right ventricle (RV), right atrium (RA), and fetal heart (Fetal). Circle 8 shows the GWAS Manhattan plot from PLINK³⁴ (circles) where only SNVs with p < 0.05 are displayed. The 283 lo-siRF-prioritized SNPs and their LD-linked (r² > 0.6) SNVs are color-coded as a function of their maximum r² value. A portion of these LD-linked SNPs (the outer heatmap layer in circle 8) are extracted from the selected FUMA reference panel (thereby with no GWAS p-values). SNVs that are not in LD (r² ≤ 0.6) with any of the 283 lo-siRF-prioritized SNVs are gray. Dashed line indicates GWAS p = 5E-8. Circle 9 shows the 21 protein-coding genes mapped by FUMA. b, Pie charts showing ANNOVAR enrichment performed for each of the 6 lo-siRF loci (circle 2 in Fig. 3a and Fig. 2f). The arc length of each slice indicates the proportion of SNVs with a specific functional annotation. The radius of each slice indicates log₂(E + 1), where E is the enrichment score computed as (proportion of SNVs with an annotation for a given locus)/(proportion of SNVs with an annotation relative to all available SNVs in the FUMA reference panel). The dashed circle indicates E = 1 (no enrichment). Asterisks indicate two-sided p-values of Fisher’s exact tests for the enrichment of each annotation. Details can be found in Extended Data 4 and 5.