RNA-Seq spliced aligners. Exon first vs Seed extend

[bob-loblaw]

Hi guys.

I was just wondering about some peoples opinions on RNA-Seq aligning methods. Maybe I'm misunderstanding it but I would think that seed extend aligners like STAR are inherently better, since they are able to make a decision on if something is spliced or unspliced. (And have the added benefit of being more suitable for mapping to a combination of human and bacteria, if one was looking at a host-pathogen interaction or something). As opposed to exon first methods (like Tophat) which map reads continuously first then deal with the spliced alignments after (which I think would bias the reads to map continuously).

What do you think?

Here is a good nature methods review discussing it for those that are interested but don't know about aligners. http://www.nature.com/nmeth/journal/...d=NMETH-201106

[dpryan]

I tend to agree. You might want to read through this paper, also from Nature Methods, that came to similar conclusions.