unsupported score in MZid af1234  2021-03-14

As previous posts in the forum I am trying to build a library from mzID files.
Specifically from this publication https://pubs.acs.org/doi/10.1021/acs.jproteome.1c00048 (mzIDs accessible here https://massive.ucsd.edu/ProteoSAFe/dataset.jsp?task=a70c9fb22a054cda94cd6d6922065701).
The files have been searched with by COMET and/or X!Tandem, and filtered with an iProphet score of ≥0.9.
Build library fails with unsupported score and when looking one of the files seems there are 'normal' comet scores like xcorr and so on.
Should I convert it to another format?


Nick Shulman responded:  2021-03-14
This is the page that lists all of the peptide search results which BiblioSpec understands:

BiblioSpec always needs to find some sort of false discovery rate value, so the search results normally need to have been run through something like Percolator or Peptide Prophet.

Those .mzid files on that FTP server certainly mention "peptideprophet" in the <AnalysisSoftwareList> at the beginning of the file, but I don't see anything that looks like a Peptide Prophet score on the SpectrumIdentificationItem's.

I am not an expert on BiblioSpec peptide search result formats. If you wait until Monday someone who is more of an expert might be able to give you better information.

Hope this helps,
-- Nick
matt.chambers42 responded:  2021-03-15
It looks like the idconvert conversion from pepXML to mzIdentML possibly stripped the PepProphet/iProphet scores, but I'm not sure how because idconvert will put in userParams if it can't convert the pepXML scores to a controlled vocabulary term. Unfortunately the repo doesn't have the original pepXMLs. It might be possible to convert back to pepXML and run through PepProphet/iProphet again but I wouldn't be surprised if that failed. The dataset was uploaded back in 2017 so the original pepXMLs might not be available anymore. Somebody could email the authors?