help with differential gene expression with cufflinks and tophat

[waterboy]

Hi All,
We are looking for differential gene expression of two different samples of the same species.

After running tophat, for two different samples of the same species, we got accepted_hits.bam files for both. Then we ran cufflinks with the following command for both of the samples:

$./cufflinks -G refGene.gtf accepted_hits1.bam
$./cufflinks -G refGene.gtf accepted_hits2.bam
where refGene.gtf is the main gtf file of the species downloaded from UCSC.

This provided us with three files, genes.expr, transcripts.expr, transcripts.gtf (for two samples, all in all, we got 6 files, genes1.expr, transcripts1.expr, transcripts1.gtf, genes2.expr, transcripts2.expr, transcripts2.gtf)

Then, we carried out cuffcompare with the following command but we are not quite sure as to this is the right way to do it or not.

$./cuffcompare -r refGene.gtf /sample1/transcripts1.gtf /sample2/transcripts2.gtf (Is this how it should be?)

Output: stdout.combined.gtf, stdout.loci, stdout.tracking

Then we carried out cuffdiff with the following command but we are not quite sure as to this is the right way to do it or not.

$./cuffdiff stdout.combined.gtf accepted_hits1.bam accepted_hits2.bam (Is this how it should be?)


Now we want to do the differential gene expression for which we guess we need to use cuffcompare and cuffdiff, right? It will be highly appreciated if anyone can guide us on the same. Also, please tell us as to what files would be input for cuffcompare and cuffdiff to do the differential gene expression.

Thank you very much for the helping hand.

[adarob]

Your commands appear to be correct.