Murphy, C;
(2016)
Dissecting the genetic architecture of cardiac disorders through the use of next generation sequencing.
Doctoral thesis , UCL (University College London).
Preview |
Text
PhDThesisFInal.compressed.pdf - Accepted Version Download (3MB) | Preview |
Abstract
The overriding goal of this thesis was to further re ne our understanding of the genetic architecture of cardiomyopathies, Arrhythmogenic Right Ventricular Cardiomyopathy (ARVC) and Hypertrophic Cardiomyopathy (HCM). 407 patients with ARVC and 957 with HCM had 41 cardiomyopathy and other putative candidate genes sequenced. By comparing these cohorts against each other and against ethnicity and phenotype matched controls, insights were gained into the role of di erent types of genetic variants in these conditions. This in part involved utilising 4500 Whole Exome Sequences (WES) that are part of the UCLexomes consortium, an in-house dataset that aggregates a diverse set of studies. High throughput DNA sequencing technologies, WES or Whole Genome Sequencing (WGS) are revolutionizing the diagnosis and novel gene discovery for rare disorders. As the eld transitions from the early discovery for Mendelian and near Mendelian diseases to more complex and oligo-genic diseases, there is substantial bene t in being able to combine data across studies, performing the type of meta-analysis for cases and controls that have proven to be so successful for Genome-Wide Association Studies (GWAS). However, WGS and WES are substantially more a ected by sequencing errors and technical artefacts than genome-wide genotyping arrays. As a consequence, meta-analysis of sequence based association studies are often dominated by spurious associations, which result in technical limitations. Here, we show that it is possible to take advantage of the type of mixed models developed initially to control for population structure in GWAS studies, and apply these ideas to control for technical artefacts. In an attempt to ascertain the role of CNVs in HCM, these data were examined for the presence of rare causative CNVs. 12 CNVs were identi ed from an initial Read Depth approach. 4 of these were subsequently validated by CoNIFER, a bioinformatics method, and Array Comparative Genomic Hybridisation (aCGH): one large deletion in MYBPC3, one large deletion in PDLIM3, one duplication of the entire TNNT2 gene and one large duplication in LMNA. These results show that the role of CNVs in HCM is small and highlight the e ciency of this two step-strategy.
Archive Staff Only
 |
View Item |
