Cufflinks get stuck at a locus

[Mayflower]

Hi, guys

I am quite new for bioinformatics. Currently, I am running cufflinks, but one sample stuck at one locus for more than two weeks, while the another sample run successfully.

Here is the command line:

cufflinks -g arabidopsis.gff -p 8 --max-bundle-frags 500000 -o mock_out accepted_hits.bam > mock_cufflinks.log

The stuck information is > Processing Locus Chr5: 26973163- 26973394 [********* ] 44%

The cufflinks I used is v 2.1.1 The accepted_hits.bam is produced by the tophat. I checked the above locus in the gff file, there is no such locus in the gff file! How can? The other sample can complete only this sample stuck here.

Any idea about this? I searched in the internet, a lot of people seemed to have the similar problem, but seems without a good solution. Thank you so much!

[dpryan]

There's a rather important difference between -g and -G that you seem to be missing. With -g, you use the GTF simply as a guide, so there will be novel features called. You might try masking that region (see the -M option).