I've run cuffdiff after running cuffmerge. These are the arguments used:

cuffdiff \
  -p 2 \
  -L S1,S2 \
  -b /data/genome.fa \
  --library-type fr-firststrand \
  --compatible-hits-norm \
  --multi-read-correct \
  -M /data/RNAseq/masks/rRNA.tRNA.gtf \
  /data/RNAseq/scripts/cuffmerge_scripts/merged_asm/merged.gtf \
  -o /data/RNAseq/cuffdiff/S1_S2

The "isoform.exp" file does not contain any refseq ids (merged.gtf does) and all rows are recorded as not significant even though there are several entries with pval<<0.05 and log2 fold change >1. I suspect it could have something to do with using the masking .gtf file, so I'm running again.

Here's a line from the output:

TCONS_00009060 XLOC_004171 DUSP8 chr11:1575280-1593150 S1 S2 OK 2.87256 6.28673 1.12997 3.24631 5.00E-05 0.435333 no

Any tips would be great. Thanks,

-Jeremy