From cuffmerge to cuffdiff: What input for cuffdiff?

Question

[BioinfGuru]

I want to compare 2 conditions: controls (4 biological replicates) v treated (3 biological replicates). I have run cufflinks successfully on all 7 BAM files independently,

I then ran cuffmerge twice to produce 2 transctipt.gtf files:

1) merged_control_transcrpts.gtf

2) merged_treated_transcripts.gtf

I was expecting that the two transctipts.gtf files would then be used as input (with BAM files) for cuffdiff. But, cuffdiff only allows 1 transctipts.gtf. Where have I gone wrong? Am I supposed to merge everything? Can cuffdiff tell the treated group from the control group?

Thanks in advance. Kenneth

Answer 1

[onspotproductions]

You need to merge all your cufflinks assemblies into a single gtf file using cuffmerge. Both conditions will be contained in a single file. You can then use this within cuffdiff and your original BAM files to test for your conditions.

[BioinfGuru]

You sir....are a legend.