Build spectral library from MSfragger SpecLib workflow af1234  2021-04-04


I am trying to build a spectral library from the output of MSfragger (SpecLib workflow). I do have the following files

  1. pepXML for all files (both individuals and combined)
  2. speclib format output
  3. osw library

However, when I try to build a DDA-Pasef spectral library using the pepXMLs there is an error because Skyline is looking for an mgf/mzXML.
When I try to use speclib the library builds but is empty and finally when I use the .tsv (which works in openswath) seems a lot of sequences are not recognized (see screenshot).
I was wondering which one would be the easiest way to build a library from these files and if any changes/conversion needs to be done to the files prior to importing into skyline.

Brendan MacLean responded:  2021-04-04

Each one of these will take some work to sort out. It seems that you are using File > Import > Assay Library with the .osw file. You could also try to build a spectral library from this file. And it would help to get your files, which you can attach to this support request if they are small enough, or upload to:

Looks like we should also improve that error message to provide the text of the invalide peptide sequence to make it easier to figure out why Skyline is having difficulty. In 20.1, I had to add code to remove preceding and trailing underscores, because some Assay Libraries specified their peptide sequences like that:


If you can figure out what Skyline is using as the peptide sequence, that may also help to debug what is going on. Though, if you provide your files, we can figure that out.

Thanks for reporting your issues.