HSA: A Heuristic Splice Alignment Tool.

Background: RNA-Seq methodology is a revolutionary transcriptomics sequencing technology, which is the representative of Next generation Sequencing (NGS). With the high throughput sequencing of RNA-Seq, we can acquire much more information like differential expression and novel splice variants from deep sequence analysis and data mining. But the short read length brings a great challenge to alignment, especially when the reads span two or more exons. Methods: A two steps heuristic splice alignment tool is generated in this investigation. First, map raw reads to reference with unspliced aligner - BWA; second, split initial unmapped reads into three equal short reads (seeds), align each seed to the reference, filter hits, search possible split position of read and extend hits to a complete match. Results: Compare with other splice alignment tools like SOAPsplice and Tophat2, HSA has a better performance in call rate and efficiency, but its results do not as accurate as the other software to some extent. Conclusions: HSA is an effective spliced aligner of RNA-Seq reads mapping, which is available at https://github. com/vlcc/HSA. [ABSTRACT FROM AUTHOR]