RegTools: Integrated analysis of genomic and transcriptomic data for the discovery of splicing variants in cancer

Abstract

AbstractSomatic mutations in non-coding regions of the genome and even exonic mutations may have unidentified non-coding consequences which are often overlooked in analysis workflows. Here we present RegTools (www.regtools.org), a free, open-source software package designed to integrate analysis of somatic variant calls from genomic data with splice junctions extracted from transcriptomic data to efficiently identify variants that may cause aberrant splicing in tumors. RegTools was applied to over 9,000 tumor samples with both tumor DNA and RNA sequence data. We discovered 235,778 events where a variant significantly increased the splicing of a particular junction, across 158,200 unique variants and 131,212 unique junctions. To further characterize these somatic variants and their associated splice isoforms, we annotated them with the Variant Effect Predictor (VEP), SpliceAI, and Genotype-Tissue Expression (GTEx) junction counts and compared our results to other tools that integrate genomic and transcriptomic data. While variants associated with certain types of alternative splicing events can be identified by the aforementioned tools, the unbiased nature of Regtools has allowed us to identify novel splice variants, including previously unreported patterns of splicing disruption in cancer drivers, such as TP53, CDKN2A, and B2M, and genes not previously reported that could represent novel driver events, such RNF145.