Errors creating a spectral library and MS1 filtering

Errors creating a spectral library and MS1 filtering michael-e-wright  2021-02-23

I keep getting errors that RAW files are missing when I build a spectral library or carry out MS1 filtering. I'm processing MS data from an external source and I want to make sure the missing MS files are truly the source of the processing errors. Thanks.

Brian Pratt responded:  2021-02-23

Correct you need the actual raw mass spec files referenced in the msms.txt file. Those are what Skyline uses to find your chromatograms.

Thanks for using the Skyline support board!

Brian Pratt

Brendan MacLean responded:  2021-02-24

The error you included on page 4 of your PDF is actually telling you that Skyline would prefer to use the raw MS/MS spectra from a raw data file for your spectral library. Ideally, you would just put those raw data files in the same folder as the msms.txt file and click the Retry button. If you click the Embedded button then Skyline will instead use the MS/MS spectra which are found in the msms.txt file itself. You should be aware, however, that these are processed spectra and may not give you accurate representations of how a peptide precursor will fragment in your mass spectrometer collision cell.

Most problematic is the fact that these spectra are "de-charged" which means that peaks found that would match fragment ions with charge states greater than 1, simply get moved to the corresponding charge 1 m/z for the fragment ion. This means you will entirely use charge 2 fragment ions, which can be a big problem for a charge 3 precursor, and even charge 1 ions may be overrepresented due to the addition of signal from the corresponding charge 2 fragment ion.

We tried for years to get the MaxQuant team to give us a way to have the true raw spectra embedded in these msms.txt file, and at first they seemed positive about doing this for us, but eventually we decided it was not a good idea for Skyline to silently include these de-charged spectra in a spectal library without mentioning the issues they can cause, without expressing a strong preference for true instrument spectra, and without even offering a way to get true instrument specta, which was the state before Matt Chambers implemented the message you are seeing.

Hope this helps to clarify. I still strongly recommend that you locate the raw data files so that Skyline can find them and use raw instrument spectra in your spectral library, but if you simply click "Embedded" then now you are at least informed of the issues you may face.