Running problem with cuffdiff

[john_nl]

Hello,

I have problem with cuffdiff, where it complains about the SAM file not been sorted correctly:

Error: this SAM file doesn't appear to be correctly sorted!
current hit is at AT1G01050.1:81, last one was at AT1G01046.1:120

However, these loci are in the correct order and running cufflinks with the same SAM files does not produce any errors.

Perhaps someone else has encountered this problem before?

Any suggestions will be gratefully received.

[weasteam]

did you use the bam file generated from tophat?

I have the same problem if I use a bam file from CASAVA, but not from tophat.

[gringer]

Error: this SAM file doesn't appear to be correctly sorted!
current hit is at AT1G01050.1:81, last one was at AT1G01046.1:120

However, these loci are in the correct order and running cufflinks with the same SAM files does not produce any errors.

That depends on your idea of "correct". Strictly, the contig/chromosome order of sorted SAM or BAM files should follow the order from the header ("The order of @SQ lines defines the alignment sorting order"), so without the header information it can't be determined. I've got no idea if cufflinks looks for a strict ordering, or a dictionary/alphanumeric ordering when checking sort order in files.

Also, if the 'AT1G...:81' refer to read IDs (rather than contig IDs), then there's not enough information in what you've given to work anything out. If the 81 and 120 correspond to base positions on the same chromosome, then the sorting is incorrect by any measure.