Error generating spectral library from FragPipe in Skyline v20.1.0.31 Chinmaya k  2020-02-13
 

Hello,

I have not been able to generate a spectral library in Skyline from FragPipe/MSFragger output. Please find the attached error file. The same error has also been reported recently as well. (https://skyline.ms/announcements/home/support/thread.view?rowId=43828)

I tried to generate a spectral library separately using BlibBuid.exe and it was also ended up with the same error. Please find the attached screenshot as well.

Is there any other way to rectify this and generate a Skyline compatible spectral library?

--
Chinmaya

 
 
matt.chambers42 responded:  2020-02-18

Hi Chinmaya,

This has been fixed but a new daily release has not yet been made that includes the fix. You can get the bleeding edge automated build (BiblioSpec.zip or SkylineTester.zip) from:
https://teamcity.labkey.org/viewLog.html?buildId=lastSuccessful&buildTypeId=bt209&tab=artifacts

Just unzip the BiblioSpec into your Skyline-daily directory, for example.

 
Chinmaya k responded:  2020-02-20

Hi Matt,

I did try using the standalone bleeding-edge BiblioSpec. But, neither it is showing an error like before nor the complete .blib file.

It is because the mzML file I have used in MSFragger was exported from Proteome Discoverer v2.2?

--
Chinmaya

 
matt.chambers42 responded:  2020-02-20

What do you get if you add "-v detail" to the command?

 
Chinmaya k responded:  2020-02-20

Hi Matt,

I did try with the above command. I can see all the PSMs in my .pep.xml file and later it will start looking for the PSM with a list scan number i.e. 626 in my case and it does not move ahead from these like before (Please find the attachment).

What might be the reason?

I could not upload the MSFragger output and mzML file in the file-sharing folder in Skyline. Is there any other way to share it with you, if you want to have a look?

--
Chinmaya

 
matt.chambers42 responded:  2020-02-21

Not sure what's going wrong there. Can you upload your pep.xml and mzML to:
https://skyline.ms/project/home/support/file sharing/begin.view?

 
Chinmaya k responded:  2020-02-21

Hi Matt,

I have added the files to the file sharing folder. It is named "MSFragger_Result_Spec_Lib.rar".

Let me know if you need more info.

--
Chinmaya

 
Chinmaya k responded:  2020-02-28

Hi Matt,

Any update on the above issue?

Were you able to figure out why the skyline or BiblioSpec is showing an error?

--
Chinmaya

 
matt.chambers42 responded:  2020-02-28

Hi Chinmaya,

I just got the chance to test with your files and unfortunately this mzML file is not valid. It uses an invalid nativeID format ("scan=123 file=123") while claiming to be Thermo nativeID ("controllerType=0 controllerNumber=1 scan=123"). What's worse, mzML only represents a single run per mzML file, while I'm pretty sure your mzML (from ProteomeDiscoverer?) is an amalgamation of multiple files. Can ProteomeDiscoverer output separate mzMLs per RAW? Those might be valid. Otherwise just use msconvert to create proper mzMLs for each RAW file and search them individually with FragPipe.

What version of ProteomeDiscoverer? This should be reported as a bug to them.

 
Chinmaya k responded:  2020-03-03

Hi Matt,

The mzML file was created by exporting unassigned spectra from a database search in Proteome Discoverer v2.2. So, yes it does contain spectra from various RAW files. I can export unassigned spectra from the individual RAW file used for search. However, it would be laborious.

We are using PD v2.2.

Can you please a screenshot of invalid nativeID from mzML file and actual nativeID? so that I can share the same with Thermo PD support team.

--
Chinmaya

 
matt.chambers42 responded:  2020-03-03

All the nativeIDs are invalid (see attachment). There is not really a way to create a valid mzML from multiple Thermo RAW files using Thermo nativeIDs at the top level. The best they could do would be to use "merged=123" nativeIDs and put the actual Thermo nativeIDs in <scan> elements: http://www.peptideatlas.org/tmp/mzML1.1.0.html#scan

Each spectrum would have 1 scan; the spectrum's id would be like "merged=123" and the scan's externalSpectrumID would be the real Thermo nativeID. The scan's sourceFileRef would point at a sourceFile element for its source raw file.

 
jmeyer responded:  2020-03-10

Where is the directory that I should place the new version of bibliospec?

Thanks,
Jesse

 
matt.chambers42 responded:  2020-03-10

If you're not comfortable doing any of the below, you should probably wait for the next Skyline-daily:

You can replace the existing BlibBuild.exe and other files in your Skyline directory. Unfortunately, for the ClickOnce installs, that directory is hidden quite deep in the user profile directory. My favorite way to find stuff like that is a handy program called Everything (because Windows search is woefully inadequate and slow):
https://www.voidtools.com/
https://www.voidtools.com/Everything-1.4.1.965.x64-Setup.exe

Pop up the search window after installing that and search for "BlibBuild.exe". That should show you all the existing Skyline installations and you can sort by Date Modified to get the most recent one (might have to right click on the column headers to make the Date Modified column visible). See attached picture for what mine looks like (you shouldn't have nearly as many entries as I do because I'm on a development machine).