Tophat/cufflinks; a way to assign read to a gene

[poisson200]

Dear all,
Can tophat or cufflinks be used to simply assign a sequence read to a gene based on genome position? Or can another program do that with SAM output or anything else (e.g. PSL)?

Cufflinks reports FPKM for genes and transcripts based on the number of reads most likely mapping to them. I think it would be cool if you had something like a --read_to_gene option that reports which reads most likely belonged to which gene/transcript.

Can this be done with Tophat/cufflinks or another program?

Thanks for any advice.

J.

Update: I just found something and this thread solves it;

http://seqanswers.com/forums/showthread.php?t=4648&highlight=assign+gene+read

Actually, it would still be good if TopHat/Cufflinks could optionally output reads to transcript assignments as it takes into account paired end data for isoform assignment. Any fresh views still very much appreciated.

[sdriscoll]

since Cufflinks runs on a SAM file it seems likely that they could carry the raw read id's through the process and output a relation file connecting Cufflinks transcript IDs to the raw reads that became a part of those transcripts. it doesn't mean it would be easy to integrate since they'd probably have to add code all over the place in the main program since it would be something that's done simultaneously with what's already happening and not before or after...but doable I'm sure it is.

[poisson200]

Thanks for the comment sdriscoll. Yes, I am guessing it would not be too hard to do something, especially if you are familiar with the code; just a bit time consuming.

I would also find it useful if it could output a FASTA sequence for all the novel transcripts or the entire transcripts cufflinks predicts. However, I am guessing the authors of Tophat/cufflinks already have a big wish list.