Cufflinks deconvolution warnings and fpkm_tracking files

[gavin.oliver]

Hi all,

I am aware that deconvolution warnings in the isoforms.fpkm_tracking files generated by cufflinks mean the entry should be treated with suspicion/ignored.

However, am I correct in thinking that the FPKM value for the corresponding gene in the genes.fpkm_tracking file should still be safe to use, since the deconvolution is a transcript-level problem?

Thanks in advance.