Integration of Short- and Long-Read RNA Sequencing Enables the Discovery of Circular RNAs.

Abstract

Circular RNAs (circRNA) are associated with crucial hallmarks of tumorigenesis. Select circRNAs contain circular open reading frames (cORF) and affect tumorigenesis through encoded small peptides. However, current circRNA detection approaches bias toward using short-read RNA sequencing (RNA-seq) for detecting circRNA backsplice junctions without reliably reconstructing complete circRNA sequences, inhibiting accurate cORF prediction. To address these challenges, we performed long-read sequencing to enrich for full-length circRNAs that could serve as a guide for short-read alignment. This approach "rescued" circRNAs eluding existing tools focused on circRNA detection from short reads and enabled the development of an open-source bioinformatics workflow that characterizes and rescues circRNAs by integrating short- and long-read RNA-seq: Characterizing CircRNAs by Integrative Sequencing (CHRIS). Application of the approach to colorectal cancer cell lines and patient samples revealed 6,445 noncanonical isoforms of known circRNAs, of which, 69 were altered during cancer metastasis. Validation experiments in colorectal cancer cell lines confirmed the endogenous expression of 11 high-confidence circRNAs rescued by CHRIS. Next, proteogenomic analysis using 67,326 circRNAs detected by CHRIS and mass spectrometry data from 261 patients with colorectal cancer from the Clinical Proteomic Tumor Analysis Consortium identified 6,848 peptides encoded by circRNAs, including 994 only detectable with long-read integration and 914 potential neoantigens. Overall, this research develops an approach that can facilitate circRNA detection and provide valuable resources for future circRNA tumor biology research. A multiomic approach integrating long-read and short-read RNA sequencing enhances circRNA detection to offer biological and clinical insights into metastatic colorectal cancer. This article is part of a special series: Driving Cancer Discoveries with Computational Research, Data Science, and Machine Learning/AI .