Whole-exome capture and next-generation sequencing to discover rare variants predisposing to congenital heart disease

Abstract

Congenital heart disease (CHD) is the most common congenital malformation, affecting 8 out of 1000 lives births, yet its aetiology remains largely unresolved. The rapidly growing number of point mutations implicated in isolated CHD suggests that single mutations may contribute significantly to CHD risk. This thesis presents an investigation of the genetic underpinnings of various types of CHD following different study designs. First, I designed a new approach to variant calling which I implemented as the variant caller BAMily. My aim was to develop a method of uncovering putative variants in next-generation sequencing data, shared by a subset of individuals and absent in another subset. I tested the variant caller’s performance against other known variant callers and demonstrated that it provides comparable; and often better, results. This novel variant caller was applied to a study of 8 families in which a disease trait was segregating; along with the variant caller SAMtools, leading to the discovery of likely disease-causing variants in 5 families. Second, I studied de novo mutation in 32 sporadic cases of transposition of the great arteries (TGA) in an attempt to identify genes that, when mutated, lead to TGA. The 32 patients with TGA were sequenced with their parents; as well as one unaffected sibling. To achieve this aim, three variant callers were used: SAMtools, GATK Unified Genotyper and BAMily, the latter acting as a filter. Potential de novo variants were found in GREB1, RBP5, SNX13. Results suggested a complex genetic etiology underlying TGA. Finally, I studied a large series of cases of tetralogy of Fallot (ToF). The study involved 824 patients which ToF and a comparator set of 490 patients with neurodevelopmental disorders lifted from the UK10K project. The aim of the study was to identify genes that, when mutated, play a role in the manifestation of ToF might cluster. For this, I first categorised variants according to their potential to disrupt protein function. I then compared genes in which potentially disease-causing rare variants occurred to lists of genes previously implicated in CHD in the literature. Following this, I identified the clustering of potentially deleterious rare variants across the coding region of genes and exons in ToF patients, hypothesising that variants influencing ToF would cluster in ToF patients. This study led to the discovery of candidate variants in FLT4 and NOTCH1 for non-syndromic ToF. As with TGA, the results I have obtained suggested a complex etiology for ToF.