sequence name mismatch error in trim.seqs

Commands in mothur
1

I’m getting a mismatch error when I try to do trim.seqs using the .fasta and .qual files I received from our sequencing center. When I run summary.seqs on the .fasta file, everything works fine. When I run trim.seqs, I get the error below, then mothur hangs and I end up having to close the terminal window.

mothur > trim.seqs(fasta=GM2GI8201.sff.fa, oligos=Britton2TagsLumped.oligos, qfile=GM2GI8201.sff.fa.qual, maxambig=0, maxhomop=8, flip=T, bdiffs=1, pdiffs=2, qwindowaverage=35, qwindowsize=50, processors=2)

sequence name mismatch between fasta: GM2GI8201CC5ON and qual file: 5
sequence name mismatch between fasta: GM2GI8201EKZTJ and qual file: 7
sequence name mismatch between fasta: GM2GI8201CJX7V and qual file: 0
sequence name mismatch between fasta: GM2GI8201E0TS3 and qual file: 6
sequence name mismatch between fasta: GM2GI8201DN5QE and qual file: 0
sequence name mismatch between fasta: GM2GI8201DS8XI and qual file: 7
sequence name mismatch between fasta: GM2GI8201CFDPW and qual file: 5

I should mention that I used these same files to process my data using RDP’s pyro pipeline, and didn’t have an issue. I saw that someone previously reported a mismatch error with trim.seqs and it turned out that the sequences giving the error had a length of zero. However, this isn’t the case for me; although some of these are short, a few are decent lengths, and BLASTing them shows that they are indeed 16S. Any thoughts or help much appreciated.
Thanks,
Laura

2

Hey Laura,

Could you re-download mothur (v.1.17.3) and give that a whirl? I think this might be related to a bug we found during the workshop. Let me know how it goes…

Pat

3

Thanks Pat. I downloaded the new version and voila - it works.

4

Hi Pat,
I think this has popped up again in mothur v.1.21.0.

It seemed to be working fine in mothur v.1.20.3.

Rohan

5

Ok so I was a bit wrong about what I said.

For some reason it doesn’t work in either version of mothur on my install on Ubuntu 10.04. I used readline and the 64bit flags on compilation. It works fine unless I specify more than one processor then it spits out a bunch of mismatch errors. I’m fasta and qual file generated in mothur from Sanger-style fastq files.

For whatever reason it all works well when I try it on OSX.

Rohan