Unmerged reads from truseq stranded RNA library, reverse complement or not?

[lordjonwald]

Hi all,

I'm trying to process metatranscriptomic reads (2x75) from a nextseq500 run, the libraries were prepared using a Truseq stranded RNA kit.

It appears that many of my paired ends do not merge, which is fine, but I'd like to concatenate the unmerged and merged reads to align against some protein databases and for de novo assembly and annotation if possible,

so my question is: is it necessary to reverse complement the second paired ends from the unmerged reads before concatenating? I imagine this wouldn’t be necessary for assemblers which like paired end reads as input but what if I wanted to input the reads as one single end read file?

Additionally (if I'm reading the truseq stranded protocol correctly) it seems that it is the antisense strand which is sequenced, if that's the case, is it actually the first paired end file which would need to be reverse complemented?

I've never worked with transcriptomic data before and I'm not entirely sure what the best practices are yet, so apologies if these are silly questions,

Anyway, any help at all would be greatly appreciated

Many thanks

[Brian Bushnell]

Aligners and assemblers don't care which strand a read comes from; you don't need to reverse-complement anything in this case. If you were mapping to a genome to discover which strand transcripts came from then it could be useful, but not otherwise.

However, you'll get more accurate mapping results by running the merged reads as single-ended and the unmerged reads as pairs in a different process, then combining the outputs.

[lordjonwald]

Thanks for your reply Brian, that really helps clear things up, I'll bear that in mind if I ever get around to mapping to a genome!