Characterizing the Genome-wide Associations of Somatic Mutations and Chromatin Accessibility in Cancer
Year: 2020
Type: dissertation
Abstract: Somatic mutation rates in cancer genomes are associated with chromatin state and informative of tumor tissue of origin. However, earlier studies considered the chromatin states of cell lines and normal tissues, while those of primary tumors remain uncharacterized. We used a machine-learning approach to evaluate tumor-specific chromatin accessibility profiles as predictors of mutation rates in 2,500 whole cancer genomes with 23 million SNVs. Mutation rates were more accurately predicted by chromatin accessibility derived from tumours than healthy cells, suggesting that somatic mutagenesis is largely associated with the chromatin state of tumour cells rather than normal cells. Interestingly, melanoma mutations were better predicted by normal melanocyte chromatin accessibility, suggesting earlier mutational timing. Furthermore, chromatin state was an accurate predictor of carcinogen-induced mutation rates while the mutations of endogenous mutational processes were only weakly statistically associated. Integrative analysis of mutations and chromatin state provides us insight into tumour evolution and heterogeneity.
Source: TSpace (University of Toronto)
Author: Oliver Ocsenas
Language: English
Subfield: Cancer Research
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W3129389623