Spectra viewing problem

[Bernt]

Hi,

I would like to continue an old thread:
http://groups.google.com/group/spctools-discuss/browse_thread/thread/4fd1ce8369511c8c/5b357a153c4a1b24?lnk=gst&q=viewing+spectrum#5b357a153c4a1b24,
which has already been taken off.

My problem is similar, but I have tried my best to verify that all the
places referring to the raw data "path" are consistent, as advised by
Jimmy in the thread. What I did was: I made a Tandem DB search at one
folder level (e.g. c:/Inetpub/wwwroot/ISB/data/OC/OGEmRPTD10/P00158),
generating a set of .raw.tandem files. But later, for the reason of
trying to combine 529 mzXML files together from separate folders, I
moved all the mzXML and associated .raw.tandem files to another folder
level (c:/Inetpub/wwwroot/ISB/data/OC/OGEmRPTD10/).

I know of this spectra viewing problem if the files were relocated, so
I used a text replacing application to ensure that all .raw.tandem
and .raw.tandem.params files with the string "c:/Inetpub/wwwroot/ISB/
data/OC/OGEmRPTD10/P00158" were replaced to "c:/Inetpub/wwwroot/ISB/
data/OC/OGEmRPTD10/". Only then I generated the .raw.tandem.pep.xml
files from the modified .raw.tandem files and then the
interact.pep.xml file.

But the error still persists: command "C:\Inetpub\tpp-bin\tar.exe -tzf
c:/Inetpub/wwwroot/ISB/data/OC/OGEmRPTD10/
HUO00159S00168GL05B01.raw.tandem.pep.tgz "*HUO00159S00168GL05B01.raw.
0659.0659.3.dta" > /dev/null" failed: Operation not permitted Error -
cannot read spectrum; tried direct .dta, from mzXML/mzData and
from .tgz

Below is an example of a spectra path that I tried to track:

Spectra URL =

&Dta=c:/Inetpub/wwwroot/ISB/data/OC/OGEmRPTD10/
HUO00159S00168GL05B01.raw.tandem.pep/HUO00159S00168GL05B01.raw.
0660.0660.2.dta&

Interact.pep.xml file =

<interact_summary filename="c:/Inetpub/wwwroot/ISB/data/OC/OGEmRPTD10/
OGEmRPTD10_Tandem_interact.pep.xml" directory="C:/Inetpub/wwwroot/ISB/
data/OC/OGEmRPTD10">
<inputfile name="HUO00159S00168GL05B01.raw.tandem.pep.xml"/>
<msms_run_summary base_name="c:/Inetpub/wwwroot/ISB/data/OC/OGEmRPTD10/
HUO00159S00168GL05B01.raw.tandem.pep"
<search_summary base_name="c:/Inetpub/wwwroot/ISB/data/OC/OGEmRPTD10/
HUO00159S00168GL05B01.raw.tandem.pep"
<parameter name="spectrum, path" value="c:/Inetpub/wwwroot/ISB/data/OC/
OGEmRPTD10/HUO00159S00168GL05B01.raw.mzXML"/>

HUO00159S00168GL05B01.raw.tandem.pep.xml file =

<parameter name="spectrum, path" value="c:/Inetpub/wwwroot/ISB/data/
OC/OGEmRPTD10/HUO00159S00168GL05B01.raw.mzXML"/>


Sorry for this long and unreadable email, but I really don't know what
else I might have missed out, and I would hate to redo the Tandem
search for 529 mzXML files. Would really appreciate some suggestions
and advice. Thanks in advance!


Bernard

[David Shteynberg]

Hello Bernard,

My best guess is that the '.' in your filename before the .mzXML is
causing some parsing to break. Please change your file basename
(everything before the extension) to include only alphanumeric
characters and try again.

Thanks,
-David

[Jimmy Eng]

Bernt,

The good news is that you shouldn't need to re-run Tandem searches to
address this problem.

The easiest way that I can think of to help you is to show you the
convention of file names, base_name attributes, and paths that the tools
typically expect/generate.

directory of data:
c:/Inetpub/wwwroot/ISB/data/OC/OGemRPTD10/

file name of one example mzXML:
c:/Inetpub/wwwroot/ISB/data/OC/OGemRPTD10/HUOL05B01.mzXML

file name of Tandem search:
c:/Inetpub/wwwroot/ISB/data/OC/OGemRPTD10/HUOL05B01.xtan.xml

file name of corresponding pepXML file:
c:/Inetpub/wwwroot/ISB/data/OC/OGemRPTD10/HUOL05B01.pep.xml

msms_run_summary base_name attribute in that pepXML file:
<msms_run_summary base_name="./HUOL05B01"

msms_run_summary base_name attribute in the combined/xinteracted
interact.pep.xml file:
<msms_run_summary
base_name="c:/Inetpub/wwwroot/ISB/data/OC/OGemRPTD10/HUOL05B01"


In your example below, you seem to carry a ".raw.tandem" extension
through everything that's a big part of the problem. After all the
tools are run and the URL is generated for the spectral viewer, some
part of this extension then is incorrectly expected to be part of the
mzXML file name.

Digest the above and see if it helps. If I were you, in addition to
doing the text replacement (that I presumed you did correctly), I would
rename your Tandem search files (.raw.tandem) based on our convention
above and re-run analysis. If the spectral viewing still fails, report
back:
- directory data is in
- an example spectrum URL as you did
- full path name of mzXML file for that example spectrum
- msms_run_summary base_name attribute value
- full path name of the pep.xml file

And instead of "HUOL05B01.*" in my example above, it's perfectly fine to
have "HUOL05B01.raw.*" if that's how you have things setup. Just as
long as everything is consistent. To follow-up on David's reply, I
believe periods (.) in filenames are fine.

Good luck.

- Jimmy

[Bernt]

Hi Jimmy, David,

Thanks for the advice, I tried removing the .raw and .d from the
extensions and changed all occurrences of the extensions in
the .tandem files accordingly. It worked. They must be interfering
with the file string recognition.

Yet when I tried the same thing for Mascot .dat files converted to
pep.xml (I removed the .mgf extension and even the ".d" hex string in
the .dat files) but it doesn't work (the url points to the
generated .tgz file, but I can't view it). I have gotten rid of all
symbols in the file names including "_" and then manually traced the
connections all the way to the mzXML files, and everything seems in
order. The last thing i can think of is the length of the filename
(Mascot search generates an index no. (e.g F065672 and attaches to the
filename). Do you think there might be something else I have missed
out?


Bernt

> >http://groups.google.com/group/spctools-discuss/browse_thread/thread/...,

[Jimmy Eng]

Bernt,

If the index number is attached to the filename and somehow carried
through to the pep.xml file and URL link, that would be a problem. Make
sure that's not the case.

The data to generate the spectral display typically comes from the mzXML
so just make sure the base_names in the resulting pep.xml files match up
to the full path mzXML file names (without the .mzXML extension).

If you checked everything and the appended index number somehow isn't
part of the spectral display URL, reply with the value of one URL and
list the full path name of the corresponding mzXML file. For the URL,
the important bit is just the value of Dta="???".

- Jimmy

[Bernt]

Hi Jimmy,

You are right, the appended index no. is indeed interfering. It's now
working and i think all the associated files
(.mzXML ; .mgf ; .dat ; .tgz ; .pep.xml) should all have the same
filenames. It's not sufficient just to ensure that the filepaths are
consistent. I don't know how the tgz file works but apparently, when i
renamed the .tgz and .pep.xml files to the same filenames as the mgf
and mzXML and .dat files, everything clicked.

Thanks a million for your help!


Bernt