Guided reconstruction of full-length isoforms from short reads by CIDANE

Abstract

Alternative splicing allows genes to express isoforms with different coding or regulatory functions on demand. While short read deep sequencing technologies (RNA-seq) provide an immediate measurement of local splicing events, the phasing of these events along full-length isoforms requires the computational inference of long-range dependencies from short-range data points. We introduce CIDANE, a tool for the assembly and quantification of full-length isoforms from short read RNA-seq data. CIDANE bridges the gap between RNA quantification methods that rely on a complete annotation of a species’ transcriptome, and transcript assembly methods that will detect novel isoforms at the cost of a lower accuracy.