multiple genes per locus for an XLOC id???

[rachelraynes]

Hello,

I am using Cuffdiff to analyze C. elegans RNA seq results. I finally made it through and upon analysis have found that there are a few occasions when an XLOC id represents a locus that includes 2 or more different genes. How do I resolve this problem? Is this a problem in the code or do I include both genes in my gene list?

Thanks!
Rachel

[peraugym]

I'd be interested in an answer as I encounter the same problem in my cuffdiff/cuffnorm output files.

Cheers,
Stefan

[dpryan]

Usually this occurs at the cufflinks or cuffmerge step. If two genes are near each other and there are enough alignments between them then they sometimes get merged. The can also happen with overlapping genes on opposite strands if you have an unstranded dataset. You can always modify the merged gtf or simply not have cufflinks look for novel features.

[bastianwur]

Well...with C.elegans you'll have polycistronic mRNA, right? -> there shouldn't be a problem with that merging.

[dpryan]

@bastianwur: Excellent point.