FFPE Aware aligner

[methylnick]

Hello fellow seq-uers, long time forum user, first time poster almost.

So I have been going through a thought experiment and we have collected transcriptome data from FFPE tumour samples. Many say, why, oh why go through the pain?

Well we have and experienced rather rubbish mapping efficiencies using bwa. Now the thought experiment, we all know of the chemical damage FFPE can do to DNA (and RNA?) and the current aligners don't take this into account, which is what I am thinking is the reason for the poor alignment.

I am wondering are there any aligners out there that can account for the known damage for FFPE? There is an analogous tool that works for bisulfite sequencing in Bismark, that uses bowtie under the hood, but accounts for the known conversion of C to T.

As far as I can tell, there isn't anything unless there is someone out there that knows of a tool?

Cheers

Nick

[Brian Bushnell]

I've been planning to put methylation/bisulfite support into BBMap for a while, but have only half-implemented it so far. Nonetheless, I suggest you give it a try; it can accommodate reads that are highly divergent from the reference, particularly if you tweak the settings for high sensitivity (for example, by reducing minid and/or decreasing kmer length). But even the default settings allow around 75% identity to the reference, and still detect long indels that are often present in tumors, or map the reads directly to the genome (as it's an RNA-seq-capable aligner).

[methylnick]

great thanks for that Brian, I don't seem to be able to find an associated publication with the aligner. It is certainly an aligner that I have not bee aware of, thanks for that.