Sequence homology in circular RNA detection

Abstract

Over the past two decades, researchers have shown an increasing interest in a special form of alternative splicing (AS) that produces a circular form of RNA distinct from the canonical linear form of RNA. Although several circular RNA detection tools have been developed, achieving both high sensitivity and high precision has been quite challenging in this area. Homologous coding sequences (exons) in the same transcript can lead to incorrect assignment of a read to a back-splicing junction instead of a linear-splicing junction, producing a source of false-positives in circular RNA detection. Although this problem has been mentioned in previous research articles, there has been no effort made to better understand the extent to which it affects sensitivity and precision. In this paper, we investigate the frequency of exon sequence homology in three different species (human, mouse and rat) and how this issue affects accurate detection of circular RNAs from high-throughput sequencing data.