Abstract:BackgroundDiscovery of cancer drivers is a major goal of cancer research. Driver genes and pathways are often predicted using mutation frequency, assuming that statistically significant recurrence of specific somatic mutations across independent samples indicates their importance in cancer. However, many mutations, including known cancer drivers, are not observed at high frequency. Fortunately, abundant information is available about functional “active sites” in proteins that can be integrated with mutations to predict cancer driver genes, even based on low frequency mutations. Further, considering active site information predicts detailed biochemical mechanisms impacted by the mutations. Post-translational modifications (PTMs) are active sites that are regulatory switches in proteins and pathways. We analyzed acetylation and ubiquitination, two important PTM types often involved in chromatin organization and protein degradation, to find proteins that are significantly affected by tumor somatic mutations.MethodsWe performed computational analyses of acetylation and ubiquitination sites in a pan-cancer dataset of 3200 tumor samples from The Cancer Genome Atlas (TCGA). These analyses were targeted at different levels of biological organization including individual genes, pathway annotated gene sets, and protein-protein interaction networks.ResultsAcetylation and ubiquitination site mutations are enriched in cancer with significantly stronger evolutionary conservation and accumulation in protein domains. Gene-focused analysis with the ActiveDriver method reveals significant co-occurrences of acetylation and ubiquitination PTMs and mutation hotspots in known oncoproteins (TP53, AKT1, IDH1) and highlights candidate cancer driver genes with PTM-related mechanisms (e.g. several histone proteins and the splicing factor SF3B1). Pathway analysis shows that PTM mutations in acetylation and ubiquitination sites accumulate in cancer-related processes such as cell cycle, apoptosis, chromatin regulation, and metabolism. Integrated mutation analysis of clinical information and protein interaction networks suggests that many PTM-specific mutations associate with decreased patient survival.ConclusionsMutation analysis of acetylation and ubiquitination PTM sites reveals their importance in cancer. As PTM networks are increasingly mapped and related enzymes are often druggable, deeper investigation of specific associated mutations may lead to the discovery of treatment-relevant cellular mechanisms.

Protein Domain-Level Landscape of Cancer-Type-Specific Somatic Mutations

Mutational landscape and significance across 12 major cancer types

Pan-cancer functional analysis of somatic mutations in G protein-coupled receptors

Analysis of somatic mutations in membrane proteins of tumor cells

Spatially Interacting Phosphorylation Sites and Mutations in Cancer

Protein structural context of cancer mutations reveals molecular mechanisms and identifies novel candidate driver genes

MutationAligner: a resource of recurrent mutation hotspots in protein domains in cancer

A full-proteome, interaction-specific characterization of mutational hotspots across human cancers

Inferring the Functional Effects of Mutation Through Clusters of Mutations in Homologous Proteins.

Fundamental immune-oncogenicity trade-offs define driver mutation fitness

A CATH domain functional family based approach to identify putative cancer driver genes and driver mutations

Somatic mutation distributions in cancer genomes vary with three-dimensional chromatin structure

Characterization of tumor suppressors and oncogenes evaluated from TCGA cancers

Frequent mutations in acetylation and ubiquitination sites suggest novel driver mechanisms of cancer

Remarkable Difference of Somatic Mutation Patterns Between Oncogenes and Tumor Suppressor Genes.

Leveraging protein dynamics to identify cancer mutational hotspots using 3D structures

Genetic Determinants of Somatic Selection of Mutational Processes in 3,566 Human Cancers

The Immune Landscape of Cancer

Interpretation of cancer mutations using a multiscale map of protein systems

Melanoma-specific mutation hotspots in distal, non-coding, promoter-interacting regions implicate novel candidate driver genes

Machine learning enables pan-cancer identification of mutational hotspots at persistent CTCF binding sites