Mascot upload mmetod  2007-02-07 04:05
Status: Closed
Assigned To: adam
 
We upgraded to 1.7 and try to use mascot to search. First we had the problems already described in other messages (prevaa etc. I recompiled mascot2xml and this was fixed. Now we have another problem ("Failed to drop temorary table". We run CPAS on a linux server and use postgres. I am attaching the log file.
Kind regards,

Metodi
 
 
jeckels responded:  2007-02-07 10:59
CPAS is failing to load this run because it's there are some protein names in the ProteinProphet results that aren't loaded into the CPAS database. There are two potential causes for this, the second of which seems more likely in your case:

1. You loaded the FASTA file with an older version of CPAS which had a bug that prevented it from fully loading all the sequence values correctly. If this is the case, you can go to Manage Site->Admin Console->protein databases, find the relevant FASTA in the top list, check its checkbox, and click on Update SeqIds. This will reload the FASTA file, ensuring that SeqIds are populated for all the proteins in the file.

2. The FASTA file that CPAS is loading doesn't completely match the FASTA file that was used for the search. If you look in the log file you posted, starting at line 639 there are a number of warnings about proteins that we're found in the FASTA file. I'm not very familiar with the Mascot2XML program that output those messages, but it seems likely that the actual search was done against a FASTA file that had those proteins, but the FASTA file available to Mascot2XML and CPAS doesn't contain them. If that's the case, you'll need to make sure that you're using the same file on both sides. Chee-Hong may be able to give some pointers here.
 
wongch responded:  2007-02-08 11:00
Hi Metodi,

Thanks for trying out the Mascot support and sorry for the inconveniences.

I will like to suggest that you try out option 1 "Update SeqIds" first. If this does not work, I will like to suggest that you create a new project in CPAS and repeat the search once more. We have not run into this situation before.

On point 2, I have added these warning as we have folded the process of looking up the protein description into Mascot2XML. (This is done via RefreshParser earlier in TPP.) These warnings do not affect the import process. However, I will like investigate on these high number of warnings. Can you zip up the folder /home/metodi/pipeline/u937/6-02-07/mascot/mascot3/ and post it here? (You can also send them to my mailbox.)

*Chee-Hong.
 
wongch responded:  2007-04-12 07:27
Based on update from Metodi's e-mail (13th Feb 2007), the issue has been resolved by changes that he made to his TPP installation.

*Chee-Hong.