Identification of germline variants that predispose to familial melanoma

Abstract

Melanoma is an extremely aggressive malignancy with a poor prognosis in advanced disease. While GWAS and exome analysis have helped to identify loci linked to the development of the disease, these studies have explained predisposition to melanoma in only a fraction of cases. Thus, the majority of the genetic factors that contribute to the pathogenesis of melanoma are yet to be defined. This project aims at identifying novel genes and pathways involved in the development of familial melanoma, and also identify loci which predispose individuals to disease development. 308 individuals from 133 different families previously diagnosed with melanoma were sequenced through a mixture of exome or whole genome sequencing. Multiple workflows were established to analyse the dataset for novel driver mutations. A novel approach of combining association and linkage analysis was established for the variants in the coding region to identify genes with high burden of mutations where the variants segregated with the disease within the pedigrees. The role of non-coding variants and structural variants in melanoma onset was also investigated through additional workflows in the whole-genome sequenced individuals. Non-synonymous mutations were found in CDKN2A, BRCA1, POT1 and BAP1. Disruptive variants were also observed in novel genes such as EXO5, TP53AIP and AMER1. An increased burden on variants in transcription factor binding motifs were observed in genes including SYK and SRC. A large deletion upstream of CDKN2A was identified. Genes including ATR and FAT1 were identified to have a higher burden of disruptive variants that segregated with the disease within the cases through the novel combined association-linkage analysis. Disruptive germline variants that could play a role in familial melanoma development were identified in multiple genes through a combination of several approaches.