Libra setup for cys-TMT labeling

CPAS Forum (Inactive)
Libra setup for cys-TMT labeling dwang7  2013-07-06 22:45
Status: Closed
 
Hi there,
I am pretty new to TMT quantitation. Apologies if my questions are too naive.

I am recently tackling iodoTMT6plex labeling on Cys. Libra algorithm, originally designed for itraq labelled at N term and Lys, is claimed to be also suitable for TMT on TPP wiki page. So I assume that labkey would also be capable of handling my situation. However, there is limited introduction to the Libra, and the main one:
http://sashimi.svn.sourceforge.net/viewvc/sashimi/trunk/trans_proteomic_pipeline/src/Quantitation/Libra/docs/libra_info.html
is pretty out of date. I generated the condition.xml file as attached file pretty much as indicated this page. My questions are:

1. can anybody help me look into the condition.xml file to see if there is any problem?
2. the default condition.xml is for itraq quantitation. So even I intended for TMT purpose, the labkey still displays something like:

iTRAQ Quantitation
Normalized 1    1.000    Absolute Intensity 1    666731.188
Normalized 2    2.782    Absolute Intensity 2    1854681.250
Normalized 3    6.175    Absolute Intensity 3    4116809.000
Normalized 4    12.833    Absolute Intensity 4    8556433.000
Normalized 5    25.960    Absolute Intensity 5    17308480.000
Normalized 6    51.994    Absolute Intensity 6    34665692.000

This is confusing since it still says "iTRAQ Quantitation", which lead me think whether Libra is working properly for my TMT.

3. TPP wiki pages says:
"If you would like to generate a condition.xml file, please use http://db.systemsbiology.net/webapps/conditionFileApp/". But the link is dead already. Can anybody update the link possibly?

4. Our TMT is labeled on Cys; so for a lot of identified peptides, if they don't have a cys, then it shouldn't display the quant information for that peptide. I guess I have to program to get rid of these fake quant information and re-assemble the protein quantitation. With that being said, Is there automatic way in labkey to automate this process? Is Libra @ TPP 4.5 suitable for my case?

5. I built TPP 4.5 rev2 in my redhat el6 linux box, together with labkey 13.1 binaries, tomcat 6, postgresql 9.2, and proteowizard 3.0. When it runs ms2, it automatically assign "tandem.exe input.xml" for searches, and of course it stops by complaining "cannot find tandm.exe". So I did the dirty change by "tandem" -> "tandem.exe", now everything works just fine. Maybe you wanna fix it in next release.

Thanks for any input in this thread! You guys are awesome to make beautiful stuff happening.

-Dong Wang
 
 
jeckels responded:  2013-07-08 15:30
Hello,

1 and 3: As far as the specifics of Libra and its condition.xml, I'd suggest that you send your questions to the TPP experts here:

https://groups.google.com/forum/#!forum/spctools-discuss

2: On the LabKey side of things, we generically (and, it seems in some cases, inaccurately) refer to Libra quantitation results as "iTRAQ", even if they're from other schemes like TMT.

4: We don't have a built-in mechanism for this. You could potentially write a custom SQL query over the peptide quantitation data (in the "peptidesdata" query in the "ms2" schema when you go to Admin->Developer Tools->Schema Browser) that automates this based on your requirements.

5: I'll open a bug to track this issue. The TPP build has historically named the XTandem binary tandem.exe, even on non-Windows systems, so I'm glad that's been changed on their side.

Thanks,
Josh
 
ruslannt responded:  2014-08-26 05:53
Bug with rename tandem > tandem.exe is not fixed in labkey 14.2 :(
 
jeckels responded:  2014-08-27 17:31
Thanks for the followup. I've identified this as a regression in 14.2. I've reopened the bug and have a fix which will be applied in 14.3:

https://www.labkey.org/issues/home/Developer/issues/details.view?issueId=18203

Thanks,
Josh