Abstract 3572: Characterization of chromosomal allelic imbalances through RNA-seq

Abstract

Abstract Transcriptome sequencing (mRNA-seq) is becoming a very versatile technique for profiling tumors, extending beyond its original intent of transcript quantification, identification of alternative transcripts, and detection of gene fusions. For instance, through recent advancements in RNA-seq data analyses, one can now computationally assess allele-specific gene expression and generate profiles of expressed somatic mutations. Here, we demonstrate the ability to identify chromosomal allelic imbalances (AI) through detection of haplotype-specific patterns in gene transcripts. This class of RNA-based observations may potentially reveal DNA-level chromosomal allelic imbalances or uncover large regions of transcription deregulation. From the TCGA Uterine Carcinosarcoma project, we downloaded exome and RNA sequencing data for 48 patients’ tumor/normal sample pairs in addition to their clinical annotations. We also downloaded Affymetrix SNP6-based DNA copy number event calls made by the TCGA for use as a gold standard when evaluating the AI calls in the exome and RNA-seq. AI calls were made in both the exome and RNA-seq data by: (1) calling 1000 Genomes genotypes in the sequencing data, (2) phasing haplotypes and then (3) characterizing haplotype imbalances using a tool that we developed called hapLOHseq. hapLOHseq applied to the exome and RNA-seq data both resulted in a 72% specificity for identifying the gold standard AI events. In RNA-seq data we detected 43% of the chromosomal AI events identified in the exome sequencing data. When considering AI events specifically detected in the RNA-seq and not the gold standard (RNA-specific AI), the data suggest that higher RNA-specific AI loads could be negatively associated with survival (p-val = 0.076), with higher RNA-specific AI load patients having a median survival of 771 days compared to 1526 days for those patients with lower loads of RNA-specific AI. In conclusion, our results suggest that analysis of chromosomal AI in RNA-seq has equal specificity for detecting DNA-level AI when compared to exome sequencing, although at lower sensitivities. Clinically, our analyses suggest that patients with higher RNA-specific AI load may have a worse overall survival prognosis. The AI we are identifying in the RNA-seq samples may reflect large-scale transcription defects, resulting in a negative impact on the survival of patients. One possible cause of RNA-specific allelic imbalance could be the presence of cis mutations that impact a large-region of the transcription of one of the two haplotypes. Currently, we are identifying areas for improvement in our analytical methods, while interrogating and characterizing exome and RNA-seq AI in additional data sets. Citation Format: Francis A. San Lucas, Yihua Liu, Zachary Weber, Erik Ehli, Gareth Davies, Paul Scheet. Characterization of chromosomal allelic imbalances through RNA-seq [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3572. doi:10.1158/1538-7445.AM2017-3572