PERTRAN: Genome-guided RNA-seq Read Assembler

Abstract

As short RNA-seq reads become a standard, affordable input to any genome annotation project, a sensitive and accurate transcript assembler is an essential part of any gene prediction system. PERTRAN is a pipeline for assembling transcripts from RNA-seq reads which demonstrates higher sensitivity, with fewer fused exons (in most cases), and faster run times compared to other TOPHAT/CUFFLINKS and genome-guided Trinity. PERTRAN shows slightly lower specificity with increased gene fusions in some cases, discussed below. SAM files generated from PERTRAN can be used to compute expression level by cuffdiff and result is comparable to that from TOPHAT.