• Sign In
  1. Requests

Peptide Import DDA library from Fragpipe Spectrum id format and file id

support

View Request

view list print
Peptide Import DDA library from Fragpipe Spectrum id format and file id naomi osullivan33127  2025-07-02 06:30
 

Dear Skyline team,
I am trying to import peptide search results from Fragpipe outputs but am encountering a format mismatch error between files.
My results are from Fragpipe v23 and am importing into Skyline (64-bit) 25.1.0.142 (7401c644b4).

I have tried multiple times with two different files to follow the Fragpipe tutorial on importing data into Skyline. During the first step, once I select the interact files and click next, I encounter the following warning (attached 1): "[SpectrumList::find] mismatch between spectrum id format of the file (scan=1) and the looked up id (controllerType=0 controllerNumber=1 scan=726)"
This warning seems to be there for every spectrum (attached 2) and thus nothing is imported into the Skyline document. If I click more info, the first line is "System.IO.IOException: WARNING: Could not find native id or title 'controllerType=0 controllerNumber=1 scan=759' in ../250128_Hela_100ng_DDA_3ms_R1_uncalibrated.mzML."

Based on the comments in the submission 'opening.d files on skyline' https://skyline.ms/announcements/home/support/thread.view?entityId=e28a2047-ead0-103d-b4e5-22f535560118&_docid=thread%3Ae28a2047-ead0-103d-b4e5-22f535560118 containing similar mismatch issue discussed in the comments, I tried moving either the raw, uncalibrated mzml or mzml into the folder but all produce the same error.

Is this a file formatting issue from Fragpipe or a Skyline issue?
Thanks for you work,
Best,
Naomi O'Sullivan
 
 
Nick Shulman responded:  2025-07-24 21:47
Sorry for the slow response.

I have successfully created a BiblioSpec spectral library in Skyline from Fragpipe v23.1 results with timstof data.
This is the step that you are failing on when you do "File > Import > Peptide Search".
The MSFragger search in Fragpipe produced some "interact...pep.xml" files and I used those to build a spectral library in Skyline.

We believe that MSFragger searches of timstof data will work with Skyline so long as the version of MSFragger is at least 4.2.

If you use MSFragger 4.1 with timstof data then you will get the "Mismatch between spectrum id format..." error.
When you run MSFragger in Skyline using the "File > Search > Run peptide search" menu item, that, unfortunately, runs MSFragger 4.1, so it would not be expected to work with timstof data.
But, we do expect it to work with Fragpipe v23 which runs MSFragger 4.2.

If you send us your files we can try to figure out what is going wrong.
You could send us some of your .d folders as well as your FASTA file.
It would also be helpful if you could send us your .pep.xml files and .mzML files.
You can upload those files here:
https://skyline.ms/files.url

-- Nick
 
naomi osullivan33127 responded:  2025-07-28 07:53
Hi Nick,

I'm actually working on Orbitrap Astral files searched in Fragpipe in DDA+ mode but the timstof files were the only ones in the support where I found this same error. I'm working in Fragpipe 23.0 but MSFragger 4.2.
I'll send the files the raw file, mzml, fasta, pepxml and interact xml documents in a zipped file to the file exchange above.
File name: 250128_Hela_100ng_DDA_3ms_R1_skyline_import

Thank you for th ehelp,
Kind regards,
Naomi
 
Nick Shulman responded:  2025-07-28 12:57
Thank you for sending those files.
I can successfully build a library using the files with names like "250128_Hela_100ng_DDA_3ms_R1_rank1.pepXML".

If I instead try to use the files with names like "interact-250128_Hela_100ng_DDA_3ms_R1_rank1.pep.xml" then I get an error about not being able to find the files whose names end in "_uncalibrated.mzML" so I cannot reproduce the error that you are running into.

The error that you are getting is very different from the error that people with timsTof data get. The timsTof error has to do with the fact that Bruker has a very weird way of identifying and ordering their spectra. With Thermo .raw files, there really is a strict ordering to the spectra that everyone agrees on, so those spectra really can be referred to by a single integer.

If it is important to you to be able to build a spectral library from the "interact-..." files, then you will need to send us the _uncalibrated.mzML files so we can reproduce the error.
I imagine the big difference between the "interact-" pep xml files and the other ones is the way that false discovery rates are calculated.
-- Nick
 
naomi osullivan33127 responded:  2025-07-30 01:10
Hi Nick,

Thank you for taking a look. Using the .pepXML documents and not the interact- produce the same intial error for me.
I did manage to build a library using only .pepXML documents only if the uncalibrated.mzml file was not in the same folder. However, the interact- files, regardless of what other files are present in the folder always produce the error. Here FDR rates aren't particularly import so now I can build a library, it's more or less resolved!

Thanks for your help,
Naomi
 
Todd responded:  2025-08-04 13:30
Hello Nick,
I would like to revive this thread as I have encountered the same issue. Unlike the OP, having FDR filtering for the import files is required for me, so I would like to use the interact*.pep.xml files for building the library. This is the first time I am using results generated from FragPipe 23.1 for building alibrary.
I have results from two instruments and two search modes in FragPpipe. (1) QE-HF DDA data searched in DDA+ mode and (2) Astral DIA data searched by DIA mode. Both dataset result in a similar error to the OP (see attached example), and an empty blib library.

EDIT: I have uploaded my FragPipe search results and intermediate files including the _uncalibrated.mzml file, as "SpecLibBuildingErrors.zip"

Any suggestions you have are much appreciated.

Best,
Todd
 
Todd responded:  2025-08-27 07:16
Hello,

I wanted to follow-up on my last post. I wondered if anyone had a chance to reproduce this error with the data files I uploaded as "SpecLibBuildingErrors.zip".

Thanks,
Todd
 
Nick Shulman responded:  2025-08-27 11:08

Todd,

I wonder if this would work if you deleted the file "20240708_Astral_IO-25cm_PC1319-Conlon_DIA_1_uncalibrated.mzML" and instead put the file "20240708_Astral_IO-25cm_PC1319-Conlon_DIA_1.raw" there.

When I try to import your search results into Skyline, I see messages like:

WARNING: Could not find native id or title 'controllerType=0 controllerNumber=1 scan=83' in .\20240708_Astral_IO-25cm_PC1319-Conlon_DIA_1_uncalibrated.mzML.

Indeed, when I open that .mzML file in a text editor, I see that the spectra are identified by only a number like:

id="scan=83"

However, I do also see that they have another attribute on the scan:

externalSpectrumID="controllerType=0 controllerNumber=1 scan=83"

That makes me think that if BiblioSpec were looking at the .raw file instead of the .mzML file, all of the spectra would have the correct IDs.

If replacing the .mzML file with the .raw file does not work, then could you send us the .raw file?
-- Nick
 
Todd responded:  2025-08-27 11:51

Hi Nick,

If i remove the mzml file from the directory and leave the .raw, the SpecLib builder isn't able to recognize it. Specifically, when I add the interact files, it returns an error even before trying to build the library. I believe it is looking for a file that is appended with in "_uncalibrated", such as "_uncalibrated.mzml" or "_uncalibrated.raw". Though I can confirm that if I manually rename the file, as you predicted, the library is generated without issue.

In the past, I have been able to build libraries using interact pepxml files from FragPipe. I haven't had time to go back and test if FragPipe changed the references in the interact files. Is it possible that Skyline used to read in base filenames even when the "_uncalibrated" was absent?

Best,
Todd