Roland,

The problem is that, for some reason, the server isn't finding any proteins that match for one or more of the protein groups that are being exported. I've already made a change in 2.2 to show an error for just those protein groups in the export instead of completely failing to export, but I'd like to figure out why the proteins aren't matching up for this run.

Does the combined run look OK when you view it through the web site? I'd expect problematic groups to have the same issues in the web site and during export. If so, are there protein groups that aren't being displayed? If there are, you could try to change the filters to determine which protein group(s) are causing the problem, and then checking to see if there's anything unusual about the protein names in the protXML file. For example, if the results are truncated after 250 protein groups, set a filter that the group number is greater than 250. If those are OK, filter for group numbers greater than 500.

Alternatively, are the mzXML, pepXML, protXML and FASTA small enough that you could send them along so that I can try to repro the problem here?

Thanks,
Josh

Josh,

     I found a protein group, which fits your description. The display on the web interface still works, but the protein information is blank (see attached screenshot). If I try to export the group I get the same symptoms as above. I tried to reproduce the issue by importing the data again in another folder, but this time there is no problem with the group and export to excel works fine.
     Is there a way to fill in the missing protein group information in the database? Or should I just delete the problematic runs and try to import them again?

Thanks.

Roland

Yes, group 29 exhibits the problem. If that run was loaded a while ago, it's worth retrying, as we've fixed a bug or two that might cause this problem in 2.0 and 2.1.

If you want to directly edit the database, you can use SQL to fill in proteins for the problematic groups, though this will be somewhat tedious and error prone. If reloading doesn't work I can try to walk you through the process.

I'd like to find out why this group failed to load correctly. Can you go back to the prot.xml file and FASTA file and post the relevant sections? I'd need the whole <protein_group> element from the prot.xml, as well as the FASTA file lines that correspond to the proteins in that group. It's probably a very long protein name that we're not matching up correctly or something along those lines.

Thanks,
Josh

Josh,

   I put the beginning of the .prot.xml, including the the first 30 groups and the sequence corresponding to group 29 in the attached file. test.tar.gz.
   Hope it helps you diagnose.

Roland

Roland,

Can you confirm if this run was loaded with 2.1 or not? There was a bug in previous versions that would have affected this particular protein, since it is longer than 80 characters. Its fix is part of 2.1.

https://www.labkey.org/Issues/home/Developer/issues/details.view?issueId=3216

The problem is with the parsing code, so if it was loaded with a previous version, it's quite possible that deleting the run and reloading it would fix the problem with that group. If you're not sure, I'd say it's worth trying to reload it. I can try to investigate further if you still have the same problem.

Thanks,
Josh