Evaluating High‐Confidence Genes in Conotruncal Cardiac Defects by Gene Burden Analyses

Figures

• 

  Figure 1. Case–control comparison of the overall burden analysis by the mean number of rare variants and individuals with at least 1 rare protein‐altering variant in 132 congenital heart disease genes. Rare protein‐altering variants were defined as missense, nonsense, frameshift, in‐frame insertion/deletion, and essential splice site variants with allele frequencies below 0.01% in the Genome Aggregation Database (gnomAD) v2.1.1 exome database of the total population. Loss‐of‐function (LoF) variants included nonsense, frameshift, and essential splice site variants. A, Mean number of rare protein‐altering variants in individuals in the conotruncal cardiac defect (CTD) cohort compared with the controls, which demonstrates an excess of rare variants per individual. B, Odds ratios (ORs) and adjusted 95% CIs of the proportions of individuals with at least 1 rare protein‐altering variant in the cohort compared with controls, with subgroups of each variant class (missense, loss of function, and in‐frame insertion or deletion). *P<0.01 (0.05/5, Bonferroni correction). Vertical dotted lines indicate the position of OR=1. Error bars indicate the 95% CI adjusted by Bonferroni correction, equivalent to…

• 

  Figure 2. Six significant genes from the gene burden analysis of rare protein‐altering variants. The proportions of individuals with rare protein‐altering variants of the 6 genes in the cohort (solid squares) were compared with the controls (open circles). Error bars indicate 95% CIs, of which no overlapping implies a significant enrichment in the cohort‐control comparison. *P<0.000893 (0.05/56 conotruncal cardiac defect [CTD]‐related genes, Bonferroni correction) ^P<0.000658 (0.05/76 other congenital heart disease [CHD] genes, Bonferroni correction). Published clinical and experimental evidence of the significant genes for their potential association with CTD were summarized with reference to the gene curation framework of ClinGen and Walsh et al, 2017. 17 MIM indicates Mendelian Inheritance in Man.

• 

  Figure 3. Significant genes from the gene burden analysis of damaging ultrarare variant burden testing in Chinese, European, and aggregated conotruncal cardiac defects cohorts. Damaging ultrarare variants are defined as protein‐altering variants that have a Phred‐scaled combined annotation‐dependent depletion score ≥20, are absent in the gnomAD v2.1.1 exome database of the total population, and appear only once in the cohort. Odds ratios compare the proportions of individuals with a damaging ultrarare variant burden of congenital heart disease (CHD) genes in the cohort compared with controls, and the respective 95% CIs are adjusted by Bonferroni corrections, equivalent to a 99.9621% CI for matching the Bonferroni corrected P value threshold of 0.000379 (=0.05/132). Vertical dotted lines indicate the position of OR=1. *P<0.000379 (0.05/132, Bonferroni correction). CTD indicates conotruncal cardiac defects; EAS, East Asian; gnomAD, Genome Aggregation Database; NFE, non‐Finnish European; and OR, odds ratio.

• 

  Figure 4. Comparison of the distribution of damaging ultrarare variants in GATA6 and FLT4. Blue dotted line: predicted amino acid change caused by missense variants; Red solid line: predicted protein truncation caused by LoF variants. All variants listed are absent in the gnomAD exome data set, appear only once in the cohort, and have Phred‐scaled combined annotation‐dependent depletion score ≥20. A, Distribution of GATA6 variants in Chinese and European cohorts. Variants are based on RefSeq transcript NM_005257.6, and protein domains are based on UniProt Q92908. B, Distribution of FLT4 variants in Chinese and European cohorts. Variants are based on RefSeq transcript NM_182925.5, and protein domains are based on UniProt P35916. Ig indicates immunoglobulin‐like domains. Note that all loss‐of‐function variants (n=2) in the Chinese (top) cohorts are canonical splice site variants. C, Comparison of the proportion of individuals with different types of damaging ultrarare variants of FLT4 in the Chinese and European cohorts. Solid squares represent the proportion in the Chinese cohort, whereas open circles represent that in the European. *P<0.05, but not reaching the significance level cutoff…

• 

  Figure 5. Distribution of pathogenic variants, likely pathogenic variants, and variants of uncertain significance among the 8 candidate genes in ClinVar. Flows from left to middle: proportion of different pathogenic classes among the 8 genes. Flows from middle to right: proportion of the variants of the 8 genes to their associated conditions. Red: pathogenic variants. Orange: likely pathogenic variants. Yellow: variants of uncertain significance. AVSD indicates atrioventricular septal defect; CTD, conotruncal cardiac defects; NDD, neurodevelopmental disorders; PHDCHD syndrome, pancreatic hypoplasia‐diabetes‐congenital heart disease syndrome; and TAAD, thoracic aortic aneurysm and aortic dissection.