pepXML

[Antonio]

I have installed TPP v4.2 on windows XP. I have created the mzXML
datafiles from my Sequest search. Now I'm making the corresponding
pepXML files. I see that the first file is being created, it seems
that is finished but I see the following warming message at the end of
the “Output so far” field:

warning: cannot open mzXML file "c:/Inetpub/wwwroot/ISB/data/AAS/
test.mzXML" for reading MS instrument info and scan times.
And the command status window indicates that the command is still
running. Is it? I do not see the file increasing in size. I can open
the pepxml file, although IE indicates that there is an error on page.
It seems to me that the program is stuck and is not progressing to
next task: the creation of the next pepXML file. Solutions?

Thanks,

Antonio

[Natalie Tasman]

Hello Antonio,

The mzXML files should correspond to the same raw file input used for
your Sequest search. For example if starting from a Thermo .raw file,
you'd convert that mzXML, search that, and then convert the search
results to pepXML.

Note that the pepXML files refer to specific file locations, so if
you've moved your mzXML file since it was first created, the TPP might
be confused.

Hope this helps and let us know how it goes,
Natalie

[Jimmy Eng]

Antonio,

That warning about not being able to open the mzXML file is not
indicative of the underlying problem. It is truly just a warning and
the analysis process is capable of finishing correctly even if the
corresponding mzXML file is missing.

The important question that's not clear from your description below is
how did you run the Sequest search? Did you run it separately via
Bioworks or did you run the search through the TPP? Although helpful
but not required, it doesn't seem like you did the latter. Do you have
a directory (e.g. c:/Inetpub/wwwroot/ISB/data/AAS/test/) that has the
Sequest .out files in it?

- Jimmy

[Antonio]

Hi Jimmy,

I run Sequest from Bioworks on our computer cluster, and then transfer
the files to the local machine running TPP, so yeas I had the c:/
Inetpub/wwwroot/ISB/data/AAS/test/
With the out-and dta- files in it.
I have attempted to run Sequest from TPP and I had the same problem;
the dta and out files are created, but the command indicates that is
still running. The problem occurs whether I send a few files or a
single one. But I have more than one file on the queue, the command
will not proceed to the next file. I can run Sequest from Bioworks on
the local machine without any problems.

> >  Antonio- Hide quoted text -
>
> - Show quoted text -

[Jimmy Eng]

Antonio,

Can you try this. Replace "basename" with one of your RAW/mzXML file
base name (i.e. <basename>.mzXML)

Collate data as

c:/Inetpub/wwwroot/ISB/data/AAS/basename.mzXML
c:/Inetpub/wwwroot/ISB/data/AAS/basename/ contains dta/out
c:/Inetpub/wwwroot/ISB/data/AAS/sequest.params

Then in a command prompt run

cd c:/Inetpub/wwwroot/ISB/data/AAS/
Out2XML basename 1

This should create basename.pep.xml in the AAS directory. This simple
exercise is just to confirm that it runs through correctly and the
process finishes. The next thing to try is to repeat the above data
structure but try the pepXML conversion in the GUI. Hopefully this
isolates whether you're experiencing a pepXML conversion problem or a
GUI problem.

- Jimmy

[Antonio]

Hi Jimmy,

OK, done it. No problem, the command were successfull, all finished,
th eTPP window indicated that the command was sucessfull and is
finished. i can open the basename.pep on TPP and all seens OK.
What next?

Antonio

> >> - Show quoted text -- Hide quoted text -

[Jimmy Eng]

I guess the next step is to repeat the pepXML conversion from the GUI
and see if the GUI reports that the command completed successfully. If
it does, organize a couple of search directories using the same
structure and repeat the pepXML conversion. Still getting stuck or is
the processes reported to be completed?

If if looks like the first pepXML file is generated/completed but the
interface thinks the command is still running, I'm going to need Luis to
chime in on what to look for as I've exhausted the bits I have knowledge
of. So let us know what happens when you try the above.

[Antonio]

Ok Jimmy,
Thank you very much for your help.
I'll repeat the entire process from TPP and I will let you know the
result.
Regards,
Antonio

[Antonio]

Hi Jimmy,
I have repeated the process from TPP. The command gets stuck either
after sending the Sequest search or creating the pepXML file. I might
have to reinstall pearl and TPP.
Antonio

[Antonio]

Hi Jimmy,
I have donne it from TPP. the program still gets stuck. pepXML files
are created and Sequest serach is performed, but the program reports

that the command is still running.

I might have a problem with the installation. I will have to reinstall/
rebuild the computer.
Antonio

On Jul 9, 2:41 pm, Antonio <aartig...@kumc.edu> wrote:

[Jimmy Eng]

Antonio,

Hopefully Luis or someone else more familiar with the GUI will chime in
and help you debug why it appears to not recognize that the pepXML
conversions apparently have completed but it still thinks they're
running. There are log files somewhere that the GUI looks at that would
be the starting point to figuring out your mystery.

[Brian Pratt]

That can happen on long searches - Apache times out, leaving your session
stuck waiting for a message that never arrives. You can adjust your Apache
config for longer timeouts (look through this group's archives for info on
that).

The next TPP GUI release features a button I added for the user to be able
to tell the GUI "no really, it's done, let's go look at the results".

[Jimmy Eng]

It's hard to tell from the dialogue whether the timeout is occuring
during the search phase or during the pepXML conversion, which are two
different tasks invoked by the GUI. I want to think it's the latter
since Antonio mentions that "pepXML conversions apparently have
completed". And if this were the case, I'm not sure that Apache time
out is the culprit since these conversions shouldn't take forever.
Anyways, that looks like a handy feature for the next GUI release.

[Brian Pratt]

Agreed - the proper test is to invoke it from the command line and see if it
actually does complete eventually. An infinite loop bug is always a
possibility.

Brian

[Jimmy Eng]

fwiw, the command line tests were successful.

[Antonio]

Jimmy,
Just to clarify: the timeout occurs both during the Sequest search and
also during the pepXML conversion. I haven't gotten the time to
reinstall the software in this computer; I hope I can do it sometime
this week.
I'll post my findings once this is completed.
Again, many thanks you for your help,
Antonio

[Jimmy Eng]

Did you try adjusting the Apache timeout parameter as Brian suggested?

Here's a description of what that parameter controls on the webserver:
http://httpd.apache.org/docs/2.2/mod/core.html#timeout

A search for 'apache timeout' in this google group brings up this thread
with some information on how to change that parameter:
http://tinyurl.com/nrr58v


- Jimmy

[Natalie Tasman]

Also, note that if you want the TPP installation to set this for you,
you'll want to uninstall Apache first. If the TPP installer detects
that you have Apache already running, it won't change the configuration.

-Natalie