Issues to build spectral library using DIAumpire-MSFragger search result

support
Issues to build spectral library using DIAumpire-MSFragger search result weixiandeng  2021-09-30
 
Hi Skyline team,

I'm trying to use DIAumpire-MSFragger search result to build the spectral library while reported the error showing below.

I cannot use Skyline built-in DIAumpire-MSAmanda workflow either, cuz it reported another error I also pasted below.

Could you please help me troubleshoot it, let me know if you need any more information from my end.

Best,
Weixian



Error reported from building spectral library:
---------------------------
Skyline-daily
---------------------------
ERROR: No spectra were found for the new library.

Command-line: C:\Users\Weixi\AppData\Local\Apps\2.0\2DGDTT6K.HWY\AXQVBVWC.NHK\skyl..tion_e4141a2a22107248_0015.0001_a9966f7f082f651e\BlibBuild -s -A -H -o -c 0.99 -i Tri_injection_GPF_DIA -S "C:\Users\Weixi\AppData\Local\Temp\tmpE51A.tmp" "E:\Research\2021-09-24-GPF-DIA\GPF-DIA\Tri_injection_GPF_DIA.redundant.blib"
Working directory: E:\Research\2021-09-24-GPF-DIA\GPF-DIA
---------------------------
OK More Info
---------------------------
System.IO.IOException: ERROR: No spectra were found for the new library.

Command-line: C:\Users\Weixi\AppData\Local\Apps\2.0\2DGDTT6K.HWY\AXQVBVWC.NHK\skyl..tion_e4141a2a22107248_0015.0001_a9966f7f082f651e\BlibBuild -s -A -H -o -c 0.99 -i Tri_injection_GPF_DIA -S "C:\Users\Weixi\AppData\Local\Temp\tmpE51A.tmp" "E:\Research\2021-09-24-GPF-DIA\GPF-DIA\Tri_injection_GPF_DIA.redundant.blib"
Working directory: E:\Research\2021-09-24-GPF-DIA\GPF-DIA
   at pwiz.Common.SystemUtil.ProcessRunner.Run(ProcessStartInfo psi, String stdin, IProgressMonitor progress, IProgressStatus& status, TextWriter writer, ProcessPriorityClass priorityClass) in C:\proj\pwiz_x64\pwiz_tools\Shared\Common\SystemUtil\ProcessRunner.cs:line 65
   at pwiz.BiblioSpec.BlibBuild.BuildLibrary(LibraryBuildAction libraryBuildAction, IProgressMonitor progressMonitor, IProgressStatus& status, String& commandArgs, String& messageLog, String[]& ambiguous) in C:\proj\pwiz_x64\pwiz_tools\Shared\BiblioSpec\BlibBuild.cs:line 201
   at pwiz.Skyline.Model.Lib.BiblioSpecLiteBuilder.BuildLibrary(IProgressMonitor progress) in C:\proj\pwiz_x64\pwiz_tools\Skyline\Model\Lib\BiblioSpecLiteBuilder.cs:line 157
---------------------------


Error reported from direct DIA search in Skyline
---------------------------
Skyline-daily
---------------------------
Error
---------------------------
OK More Info
---------------------------
System.IO.FileNotFoundException: Could not find file 'C:\Users\Weixi\AppData\Local\Temp\4ml3jgnk.uoq.mz5'.
File name: 'C:\Users\Weixi\AppData\Local\Temp\4ml3jgnk.uoq.mz5'
   at System.IO.__Error.WinIOError(Int32 errorCode, String maybeFullPath)
   at System.IO.File.InternalMove(String sourceFileName, String destFileName, Boolean checkHost)
   at pwiz.Skyline.Model.DiaUmpireDdaConverter.Run(IProgressMonitor progressMonitor, IProgressStatus status) in C:\proj\pwiz_x64\pwiz_tools\Skyline\Model\DiaUmpireDdaConverter.cs:line 156
---------------------------
 
 
Nick Shulman responded:  2021-09-30
I am not sure what is going on with the first error you have reported, but that second error is probably because Skyline has a bug in it that the the DIAUmpire search only works on data that has already been centroided.

You should use ProteoWizard msconvert to create .mzML files, and be sure to add the "Peak Picking" filter so that the data will be centroided. (MSConvert calls it "peak picking" but Skyline calls it "centroiding").

Then, for the DIA search in Skyline, use these .mzML files instead of your raw files.

We will fix this bug in a future update of Skyline-Daily, but, for now, for the DIAumpire search, you must centroid your data first.

I am not sure what is going on with that first "No spectra were found for the new library". If you would like, you can send us your files and we can take a look.
Here is the web page that tells you what types of files you need to build a spectral library depending on the type of peptide search engine you used:
https://skyline.ms/wiki/home/software/BiblioSpec/page.view?name=BlibBuild
Sometimes people have trouble building a library because they are trying to use the incorrect type of files.

You can send us your files and we can take a look. If your files are less than 50MB you can attach them to this support request. Otherwise, you can zip them up and upload them here:
https://skyline.ms/files.url
-- Nick
 
weixiandeng responded:  2021-09-30
Wow, thanks Nick for your swift response! I will try with the centroid format.

I uploaded one of my pep.xml file, if you don't mind, please check it out.

Best,
Weixian
 
weixiandeng responded:  2021-09-30
One more question, if I have files with different isolation schemes and I want to process them with Skyline built-in DIAumpire module together, how should I get around the isolation scheme option? Choosing results only cannot help me advance to the search.

Best,
Weixian
 
Kaipo responded:  2021-09-30
Hi Weixian,
Would you mind also uploading the mzML file(s) associated with these search results?

Thanks,
Kaipo
 
weixiandeng responded:  2021-09-30
Thanks for taking look at the issue, I just uploaded the mzml file.
 
Kaipo responded:  2021-09-30
Hi Weixian,
BiblioSpec is having trouble matching the spectra in the pep.xml to the ones in the mzML file.
Is the mzML file the same one used with MSFragger to search? I'm seeing some unexpected things in these files, e.g. the spectrum names have a different format. In the pep.xml file this spectrum:
"2021-09-24-WD-140min-DIA-1cv_40V-GPF-84window-MixA_Q1.00001.00001.3"
corresponds to this one in the mzML with a slightly different name:
"2021-09-24-WD-140min-DIA-1cv_40V-GPF-84window-MixA_Q1.1.1.3"

Thanks,
Kaipo
 
weixiandeng responded:  2021-09-30
Thanks for checking into it, then this seems an issue with Percolator, probably the developer changed the spectrum title naming rule in the recent updates? Do you have any way to get around it from your end?

Best,
Weixian
 
Kaipo responded:  2021-09-30

I was able to get the library to build by adding this:

<sourceFileList count="1">
  <sourceFile id="test1" name="test.RAW" location="file://.">
    <cvParam cvRef="MS" accession="MS:1000774" name="multiple peak list nativeID format" value=""/>
  </sourceFile>
</sourceFileList>

to the <fileDescription> element in the mzML file. This causes the library builder to match spectra via index numbers, but I'm not sure if the correct spectra are being linked to each other though, because some of the spectra didn't look like they belonged to the matched peptides.
Is it possible to re-run MSFragger using a different format than mzML (mgf or mzXML)? You might have more success building the library with another format.

Thanks,
Kaipo