Subset trinity transcripts to remove contamination

[jmp92]

I have a set of trinity-assembled transcripts that are supposed to be eukaryotic, but it appears that bacterial contamination was an issue during the experiment. However, I'd still like to make the most of this data, so I ran blastx with the transcripts and made a subset of only those that had a eukaryote as a top hit.

So now I have a fasta file with a subset of trinity-assembled transcripts and I am trying to figure out how to use this to eventually look for differential expression. I am new to trinity and am wondering: Is it useful to align the original rna-seq reads to this subset? Will I have issues because the reads from the contaminant (hopefully) will not align to any transcripts in this subset?

(the mapping I'm referring to is outlines in the following: https://github.com/trinityrnaseq/tri...inity-Assembly)

Thanks!

[Brian Bushnell]

I would generally recommend determining what the bacterial contaminants are, mapping the reads to the whole genomes of the contaminant bacteria, then using the unmapped reads only for assembly and expression analysis.