Can Cuffmerge give FPKM values?

[Luyi Tian]

I am totally confused about the FPKM given by cuffmerge. This is my pipeline:
I have 6 replicates of the sample and run tophat and cufflink to each of them. Then I run cuffmerge and hope I can get gene abundance of this sample but most FPKM are one and some with extremely low value(e-324): my genes.fpkm_tracking file is like this:

Eif2s2 - - Eif2s2 Eif2s2 TSS1774 2:154697145-154718642 - - 4.94066e-324 4.94066e-324 4.94066e-324 OK
Cbfa2t2 - - Cbfa2t2 Cbfa2t2 TSS8205 2:154262219-154365092 - - 9.88131e-324 9.88131e-324 9.88131e-324 OK
CUFF.22438 - - CUFF.22438 - - 2:154299058-154299334 - - 1 1 1 OK
CUFF.22439 - - CUFF.22439 - - 2:154329343-154329686 - - 1 1 1 OK
CUFF.22440 - - CUFF.22440 - - 2:154352123-154352545 - - 1 1 1 OK

about 70%reads could be mapped to the annotation and the cufflink output of each replicates seems ok, this is the genes.fpkm_tracking file of cufflink:

CUFF.29200 - - CUFF.29200 - - 15:53120262-53120485 - - 10.2459 3.76582 16.726 OK
CUFF.29201 - - CUFF.29201 - - 15:53120618-53120878 - - 6.34698 2.3328 10.3612 OK
CUFF.29202 - - CUFF.29202 - - 15:53121871-53122016 - - 96.5221 54.3964 138.648 OK
CUFF.29204 - - CUFF.29204 - - 15:53122505-53122811 - - 7.2459 3.83015 10.6616 OK

Anyone knows if cuffmerge can give reasonable FPKM values, and if it cannnot, how to get an 'average' gene expression levels of this replicates.


Thanks in advance for advice
Luyi

[Bulak]

Any follow-up on this? Strangely I have exactly the same numbers (4.94066e-324 and 1) for almost all isoforms and genes. Luyi, did you find some answers in the meanwhile? Thanks.

Bulak