I am using tophat 1.3.0 and cufflinks 1.0.3 with -g (GTF guided) option and -u (allocate multi-reads). I am seeing a number of genes in all my samples (45 samples) that have reads aligned to them using tophat, but do not show up in the genes.fpkm_tracking result file from cufflinks. It seems to be a similar set of (~14,000) genes in all the samples that are missing, and they range all over the spectrum of read counts.
We noticed this because 3 genes of interest to the researcher were missing from the results, but they were present with FPKMs at reasonable levels when we did the analysis previously (with older versions of everything).
Do you have any suggestions for me as I try to troubleshoot this problem?
Right now, I'm trying cufflinks using a gff file (instead of gtf) and trying an older version of cufflinks to verify that recovers the genes. My gtf file has one 4 megabase long intron I found using cufflinks gffread, but I assume that wouldn't cause this type of problem.
I also emailed this to [email protected], so I'll post back if I get an answer.
We noticed this because 3 genes of interest to the researcher were missing from the results, but they were present with FPKMs at reasonable levels when we did the analysis previously (with older versions of everything).
Do you have any suggestions for me as I try to troubleshoot this problem?
Right now, I'm trying cufflinks using a gff file (instead of gtf) and trying an older version of cufflinks to verify that recovers the genes. My gtf file has one 4 megabase long intron I found using cufflinks gffread, but I assume that wouldn't cause this type of problem.
I also emailed this to [email protected], so I'll post back if I get an answer.
Comment