Identifying, Prioritizing, and Visualizing Functional Promoter SNVs with the Recurrence-agnostic REMIND-Cancer Pipeline and pSNV Hunter

Abstract

Cancer is a heterogeneous disease that arises due to mutations that drive cancer progression. However, the identification of these functional mutations has typically focused only on protein-coding DNA. Among non-coding mutations, only a few have been clearly associated with cancer. We hypothesize that this gap in discovery is partly due to the limitations of current methods requiring high recurrence of mutations. To support candidate selection for experimental validation of lowly recurrent and singleton promoter mutations, new computational approaches for the integrated analysis of multi-omics data are required. To address this challenge, the REMIND-Cancer Pipeline leverages whole-genome sequencing and RNA-Seq data to extract and prioritize functional promoter mutations, regardless of their recurrence status. Subsequently, pSNV Hunter aggregates and visualizes comprehensive information for each candidate. We demonstrate the functionality of both tools by applying it to the PCAWG dataset. This workflow successfully identified and prioritized known highly-recurrent mutations, as well as, novel singletons and lowly recurrent candidates. Hence, the output of our workflow directly supports hypothesis generation for subsequent experimental validation to overcome limitations of recurrence-based approaches.