Searching MGF-format files

33 views
Skip to first unread message

Viktor Granholm

unread,
Jan 10, 2013, 10:48:09 AM1/10/13
to crux-users
Hi group,
I have a question regarding searching spectra-files in the MGF-format, using Crux v1.39. 

I have set the --spectrum-parser option to pwiz, but I seem to get an empty output. Should Crux automatically detect that I'm using an MGF-file, or should I specify it somehow? In the standard-error, it says INFO: Reading in ms2 file CID_IT_Tryp.mgf
Which perhaps suggests that Crux thinks I'm using an ms2-file?

Thanks for any help, and tell me if more information is needed.
Viktor

Kaipo Tamura

unread,
Jan 10, 2013, 12:08:18 PM1/10/13
to crux-users
Hi Viktor,

I believe that this is a bug. Both the pwiz and crux parser options
should automatically recognize MGF files by their extensions (the
"Reading in ms2 file" message appears regardless of the filetype).

I will have a look at the code and see if I can figure out what is going
on. Something to try in the meantime would be using the default spectrum
parser (i.e. "--spectrum-parser crux") and see if the output is still empty.

Thanks,
Kaipo
> --
>
>

Kaipo Tamura

unread,
Jan 10, 2013, 2:59:33 PM1/10/13
to crux-users
Hi Viktor,

I attached a patch that should resolve the issue. If you would prefer a
binary or the patch does not work please let us know.

Thanks,
Kaipo
pwiz_parser.patch

Viktor Granholm

unread,
Jan 11, 2013, 7:25:29 AM1/11/13
to Kaipo Tamura, crux-users
Hi Kaipo,
thanks for the help. The patch seems to work just great.
Best,
Viktor


2013/1/10 Kaipo Tamura <kai...@uw.edu>


--



Viktor Granholm

unread,
Jan 18, 2013, 5:12:58 AM1/18/13
to Kaipo Tamura, crux-users
Hi again Kaipo,
it took me some time to realize it, but I've discovered a problem when running Crux v1.39 in the 'sequest-search'-mode and the pwiz_parser.patch on MGF-format files. I suppose the 'sequest-search' mode might be deprecated by now, but the problem might be seen elsewhere too. In the SQT-format output, the scan numbers (or spectrum numbers) are always 0, which makes downstream processing hard.
Is there a way to solve this, or would you like more information?
Thank you,
Viktor


2013/1/11 Viktor Granholm <viktor....@scilifelab.se>
Reply all
Reply to author
Forward
0 new messages