Naslov
Tissue-of-origin chromatin organization shapes the mutational landscape of cancer.

Autori
Polak, Paz ; Karlić, Rosa ; Koren, Amnon ; Thurman, Robert ; Sandstrom, Richard ; Reynolds, Alex ; Rynes, Eric ; Stamatoyannopoulos, John A. ; Sunyaev, Shamil R.

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Skup
63rd Annual Meeting of the American Society of Human Genetics

Mjesto i datum
Boston (MA), Sjedinjene Američke Države, 22.10.2013. - 26.10.2013

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
cancer ; mutional density ; chromatin ; epigenomics

Sažetak
Density of somatic mutations in cancers varies greatly along the genome, and chromatin organization has been implicated as a major determinant of the mutational landscape in cancer. However, both the regional variation of somatic mutations and chromatin organization are highly cell type specific properties. We analyzed genomes of five major types of cancer representing a range of tissues of origin and mutational mechanisms and signatures. We compared the genomic distribution of cancer mutations to a large set of epigenetic properties that included chromatin accessibility as measured by DNaseI hypersensitivity, histone modifications and DNA replication timing from multiple cell types. In order to evaluate the contribution of different epigenetic properties to the local variation in mutation density, we used a machine learning tool called Random Forest regression. Random Forest provides a framework to rank individual predictors and to identify the set of strongest predictors of local mutation density, and also numerically scores the contribution of each individual predictor. For each cancer type, we focused on the predominant mutation type and initially combined all mutations from the different genomes of that same cancer type. We analyzed mutation density variation at 1Mb scales. Remarkably, epigenetic marks explain a large fraction of variance in mutation density for all cancer types. Variance explained ranges from 65% to 88%. This is substantially higher than in earlier studies, and indicates that, at least for several cancer types, we have identified a set of epigenetic variables that almost fully predict the mutation variability along the genome. Furthermore, chromatin marks from the matched tissue of the cell-of-origin of a cancer, e.g. melanocytes and melanoma, explain a much larger proportion of the variance than the chromatin marks from non-matched tissues.

Izvorni jezik
Engleski

Znanstvena područja
Biologija

POVEZANOST RADA