Nour Hanafi

BACKGROUND: Congenital heart disease (CHD) is the most common birth defect. Ninety percent of isolated cases remain genotype-elusive despite short-read genome sequencing (GS). Our goal was to assess genomic yield with long-read GS compared with short-read GS in CHD. METHODS: We performed Illumina short-read GS on 1101 CHD probands. In a subset of 46 genotype-elusive probands, we performed additional PacBio long-read GS as well as parental sequencing of 7 complete trios. We compared variant calls genome-wide, including across dark and CHD genes in samples with paired short-read and long-read GS. RESULTS: Paired analysis of the 46 probands revealed that genome-wide, long-read GS had 1.01- to 7.93-fold higher call rates of single nucleotide variants, deletions, duplications, and insertions but fewer indels and inversions compared with short-read GS. With a genome-wide sequencing depth of 18.3×, long-read GS had higher coverage for 10 Tier 1 CHD genes and dark genes, but nonuniform and low sequencing depth (ZEB2 in a CHD proband who had an extracardiac phenotype overlapping with Mowat-Wilson syndrome. CONCLUSIONS: Long-read GS demonstrated higher genome-wide variant call yield compared with short-read GS, better coverage of several cardiac-relevant genes, and better resolution of complex structural variants, including tandem repeat expansions. Long-read GS may provide an option for a subset of patients with CHD who remain genotype-elusive on short-read GS.

Congenital heart disease (CHD) is the most common birth defect. We performed Illumina short-read genome sequencing (GS) of 1,101 probands, which identified a genetic cause in 16% of cases. We performed PacBio long-read GS in 43 genotype-elusive patients. Paired analysis revealed higher detection with long-read GS of single nucleotide variants, deletions, duplications, and insertions as well as fewer false-positives for indels and inversions. Long-read GS had higher coverage of nine CHD genes, but a low sequencing depth (de novo structural variant upstream of ZEB2 that was only resolved with long-read GS in a patient with extra-cardiac phenotype overlapping Mowat-Wilson syndrome. Long-read GS may provide an option in CHD patients who remain genotype-elusive on short-read GS.

Congenital heart disease (CHD) is the most common congenital anomaly. Non-canonical splice-disrupting variants are not routinely evaluated by clinical tests. Algorithms including SpliceAI predict such variants, but are not specific to cardiac-expressed genes. Whole genome (WGS) (n=1083) and myocardial RNA-Sequencing (RNA-Seq) (n=114) of CHD cases was used to identify splice-disrupting variants. Using features of variants confirmed to affect splicing in myocardial RNA, we trained a machine learning model that outperformed SpliceAI for predicting cardiac-specific splice-disrupting variants (AUC 0.92 vs 0.66), and was independently validated in 43 cardiomyopathy probands (AUC 0.88 vs 0.64). Application of this model to 971 CHD WGS samples identified 9% patients with splice-disrupting variants in CHD genes. Forty-one% of predicted splice-disrupting variants were deeply intronic. The burden of variants in CHD genes was higher in cases compared with 2,570 controls. Our model improved genetic yield by identifying splice-disrupting variants that are not evaluated by routine tests.

Genome editing with CRISPR-associated (Cas) proteins holds exceptional promise for “correcting” variants causing genetic disease. To realize this promise, off-target genomic changes cannot occur during the editing process. Here, we use whole genome sequencing to compare the genomes of 50 Cas9-edited founder mice to 28 untreated control mice to assess the occurrence of S. pyogenes Cas9-induced off-target mutagenesis. Computational analysis of whole-genome sequencing data detects 26 unique sequence variants at 23 predicted off-target sites for 18/163 guides used. While computationally detected variants are identified in 30% (15/50) of Cas9 gene-edited founder animals, only 38% (10/26) of the variants in 8/15 founders validate by Sanger sequencing. In vitro assays for Cas9 off-target activity identify only two unpredicted off-target sites present in genome sequencing data. In total, only 4.9% (8/163) of guides tested have detectable off-target activity, a rate of 0.2 Cas9 off-target mutations per founder analyzed. In comparison, we observe ~1,100 unique variants in each mouse regardless of genome exposure to Cas9 indicating off-target variants comprise a small fraction of genetic heterogeneity in Cas9-edited mice. These findings will inform future design and use of Cas9-edited animal models as well as provide context for evaluating off-target potential in genetically diverse patient populations.