"I’m running cufflinks on a scaffolded genome. Therefore some transcripts are actually split by the ends of scaffolds. Cufflinks reports a few transcripts as having ends greater than the lengths of the scaffolds, I imagine because the start plus the length of the aligned read goes off the end. This causes the PASA pipeline to crash, and can cause other such scripts to fail. As the BAM file contains the lengths of each scaffold, would it be possible to enforce that this doesn’t happen?"
From Bob Zimmerman:
"I’m running cufflinks on a scaffolded genome. Therefore some transcripts are actually split by the ends of scaffolds. Cufflinks reports a few transcripts as having ends greater than the lengths of the scaffolds, I imagine because the start plus the length of the aligned read goes off the end. This causes the PASA pipeline to crash, and can cause other such scripts to fail. As the BAM file contains the lengths of each scaffold, would it be possible to enforce that this doesn’t happen?"