crux-1.39 is not working

[Taejoon Kwon]

Hi, 

I've tried to run crux-1.39 on my ubuntu machine (10.10) but I failed without any messages. (1.38 version is working fine)

Here's messages I've got with -verbosity 100 option. Any idea?

----

DETAILED DEBUG: Checking option 'use-flanking-peaks' of type 'bool' for type and bounds

DETAILED DEBUG: found boolean_type param, value 'false'

INFO: Beginning search-for-matches.

WARNING: The output directory '.' already exists.

Existing files will be overwritten.

INFO: CPU: cygnus

INFO: Fri Dec 14 12:26:09 CST 2012

DEBUG: Printing parameter file

DEBUG: Using crux to parse spectra

INFO: Reading in ms2 file ../mzXML/2012nov30orbi_PSEAEpao1_HH1a.mzXML

DEBUG: There were 0 spectra found in the ms2 file

INFO: Preparing protein fasta file ../DB/PSEAE_PAO1_combined.fa

DEBUG: Creating new database from '../DB/PSEAE_PAO1_combined.fa'

DEBUG: Parsing text fasta file '../DB/PSEAE_PAO1_combined.fa'

DEBUG: Found 11144 proteins

DEBUG: OutputFiles is opening 1 files (0 decoys) in '.' with root '2012nov30orbi_PSEAEpao1_HH1a.PSEAE_PAO1_combined'. Overwrite: 1.

DEBUG: Printing parameters to xml output

DEBUG: Queueing next spectrum

DEBUG: Done Queueing next spectrum

INFO: Created 1 peptide mods, keeping 1 with 255 or fewer aa mods

DETAILED DEBUG: Database pointer count 11144

DEBUG: Freeing database.

DEBUG: Closing database filehandle

INFO: Elapsed time: 0.58 s

INFO: Finished crux-search-for-matches

INFO: Elapsed time: 0.58 s

INFO: Finished crux search-for-matches.

INFO: Return Code:0

----

Just in case, I also posted mzXML file, FASTA file and running script at http://www.marcottelab.org/users/taejoon/crux_failed/.

Thanks, 

Taejoon

[Sean McIlwain]

Taejoon,

Could you try setting --spectrum-parser=pwiz or --spectrum-parser=mstoolkit in your search-for-matches command line?

Thanks,
Sean

--
 
 

[Charles Grant]

Hi Taejoon,

On Dec 14, 2012, at 10:44 AM, Taejoon Kwon <taejoo...@gmail.com> wrote:

> I've tried to run crux-1.39 on my ubuntu machine (10.10) but I failed without any messages. (1.38 version is working fine)
>
> Here's messages I've got with -verbosity 100 option. Any idea?
>

> INFO: Reading in ms2 file ../mzXML/2012nov30orbi_PSEAEpao1_HH1a.mzXML
> DEBUG: There were 0 spectra found in the ms2 file

It looks like Crux is having some problem reading the spectrum files. We'll see if we can reproduce this and debug the problem. I'm downloading the input files from your web page.

Thanks for letting us know about the problem! We'll try to get back to you shortly with an answer.

Charles

[Taejoon Kwon]

Dear Sean,

'mstoolkit' works, and 'pwiz' doesn't. Is crux-1.39 set to use pwiz as
default parser? I am also curious whether different sub-programs use
different parser.. I am testing Hardklor now and it seems to be
working.

thanks,

Taejoon

--
-=-=-=-=-=-=-=-
Taejoon Kwon, Ph.D <taejoo...@marcottelab.org>

Edward Marcotte Lab
Center for Systems and Synthetic Biology/Institute for Cellular &
Molecular Biology
University of Texas at Austin, USA

[Sean McIlwain]

On Fri, Dec 14, 2012 at 1:47 PM, Taejoon Kwon <taejoo...@marcottelab.org> wrote:

Dear Sean,

'mstoolkit' works, and 'pwiz' doesn't. Is crux-1.39 set to use pwiz as
default parser? I am also curious whether different sub-programs use
different parser.. I am testing Hardklor now and it seems to be
working.

That's interesting, I thought both of those would work.  Hardklor and Bullseye both use mstoolkit to parse spectra.  crux search-for-matches can either use the internal crux parser (default), mstoolkit, or proteowizard.  Could you send me the log file for crux search-for-matches when using the proteowizard parser?

Thanks,
Sean
 




 

[Taejoon Kwon]

Here you are.. Just in case, I also attached params.txt file.

Thanks,

Taejoon

[Kaipo Tamura]

Hi Taejoon,

The fix for using pwiz as the parser will be included in the next
release. In the meantime, let us know if you'd like a patch/updated
binary (if you'd prefer to use the pwiz parser instead of the mstoolkit
one).

Thanks,
Kaipo