Cufflinks: Why in the skipped gtf file showed a portion of genome masked

Question

[mjoyraj]

I mapped my RNA-se reads with Tophat with the commands

Tophat -p 8 -o output -I 20000 -g 5 -G gtf_file Genome Fastq1 fastq2

Then I assembled the aligned tophat bam file using cufflinks with the commands

Cufflinks -p 20 -o output accepted.bam file

I found some of the genes are not assembled and they are masked as can be seen in skipped.grf.

I wonder why this has happened. Can anybody suggest something

Answer 1

[Devon Ryan]

1. That tophat command is appropriate for PE mapping. You need to separate the fastq files with commas (no spaces!).
2. If coverage is too high then cufflinks will skip a region.

[mjoyraj]

Yes, I observed the skipped region are mostly my target genes which are highly expressed (as can be seen from mapped reads). So I am trying again cufflinks-assembly with "--max-bundle-frags 1000000000000".

[Devon Ryan]

Something like that, I always have to look up cufflinks parameters.

[mjoyraj]

Thanks.... Ryan

[mjoyraj]

Dear Ryan, when I used --max-bundle-frags 1000000000000 in cufflinks, the run continues for a long time and it is still running showing waiting for two threads to complete. Is it okay?. Do I need to increase the number of threads?