I was just checking the new cufflinks versions (2.0.0 and now 2.0.1) with a test dataset. The recent versions appear to take way too long to build the gtf files compared to the old version. Anyone know why?
./newcuff_ABMN_R1:
total 99528
4437383 Jun 27 14:03 genes.fpkm_tracking
5843960 Jun 27 14:03 isoforms.fpkm_tracking
0 Jun 27 09:04 skipped.gtf
91618081 Jun 27 14:03 transcripts.gtf
./oldcuff_ABMN_R1:
total 99644
4436635 Jun 26 17:20 genes.fpkm_tracking
5858163 Jun 26 17:20 isoforms.fpkm_tracking
0 Jun 26 16:36 skipped.gtf
91724175 Jun 26 17:20 transcripts.gtf
Look at the timestamps (diff between the timestamps of skipped.gtf and transcripts.gtf). the old version took about 45 mins where as the new version took 5 hours!! - both using the same accepted_hits.bam file and identical parameters.
./newcuff_ABMN_R1:
total 99528
4437383 Jun 27 14:03 genes.fpkm_tracking
5843960 Jun 27 14:03 isoforms.fpkm_tracking
0 Jun 27 09:04 skipped.gtf
91618081 Jun 27 14:03 transcripts.gtf
./oldcuff_ABMN_R1:
total 99644
4436635 Jun 26 17:20 genes.fpkm_tracking
5858163 Jun 26 17:20 isoforms.fpkm_tracking
0 Jun 26 16:36 skipped.gtf
91724175 Jun 26 17:20 transcripts.gtf
Look at the timestamps (diff between the timestamps of skipped.gtf and transcripts.gtf). the old version took about 45 mins where as the new version took 5 hours!! - both using the same accepted_hits.bam file and identical parameters.