Cufflinks output - Next wat?

[ritzriya]

Hi everyone,

I have run Cufflinks on the RNA-Seq Illumina data and it has produced .gtf file in the end.

I am visualising the gtf and sam file on the visualisation tools, but now i need to know what is the gene, it is matching to? How will i get that? Is there a tool available for this or do i need to code for it? For example, if have the regions listed thru cufflinks, shud i find a way to BLAST the result to get the names of the genes i have finally got or is there some other way?

Waiitng for a reply soon,

Thanks!

[ritzriya]

Strange case - new to cufflinks!

Also, I want to know what if the cufflinks has produced a exon aligning to the reference genome if viewed on a visualizer, donot show any reads aligning in that region, then shoould I consider that gene as my candidate gene or simply take it as a gene?

Waiting for your reply,
Ritzriya..

[GKM]

What you do next depends on what you want to know. I suggest reading the manual, specifically, the parts about cuffcompare and cuffdiff, but without knowing what the goal of your analysis is, it's hard to tell what you should do with the gtf files.

Also, I want to know what if the cufflinks has produced a exon aligning to the reference genome if viewed on a visualizer, donot show any reads aligning in that region, then shoould I consider that gene as my candidate gene or simply take it as a gene?

What kind of track are you looking at?

[ritzriya]

Hi GKM,

My goal of analyzing gtf files will be to be find novel genes in a species where it is aligned to a known nearby species. So if i see that a known gene in the other species is getting aligned by the reads produced by sequencing my species and is covering the entire portions of exons, then I would call it a gene found in my organism.

Henceforth, i require as to which gene do my reads are actually getting mapped to? I hope you got my point.

[GKM]

The reads aren't mapped to any gene though, TopHat maps to the genome.

If I understand you properly, you have two species, one annotated and one now, and you want to assemble transcripts from the latter based on RNA-Seq data and look for those that have high homology to the annotated transcripts in the other former. Then you want to take the sequence of those transcripts (you will have to assemble them on your own, I don't think cufflinks has that option right now) and BLAST them against the transcriptome of the other species. You may want to do some prevalence filtering before that though to discard the obvious garbage and to speed up things.

[ritzriya]

How will i get the transcriptome of the other known species, if i need to BLAST against them?