I suspect that all of the spectra were skipped:

WARNING: Skipped 25714 spectra with invalid precursor info

You already indicated that you suspected something wrong with the scan headers. Did you modify them in some way?

Normally standard mzML files produced by MSConvert, ThermoRawFileParser, etc. should all work. We do not edit the mzML files or the headers in there at all.

Hi @bittremieux

Yes i suspect the headders as my data orriginates from a timsTOF with the IM engaged. I don't think that the IM is to blame as it is handeled in the conversion (see command below). Given that the data is from a timsTOF I do not think the ThermoRawFileParser is used at all.

For info, the CMD command i use to generate the mzML files is something along the lines of this : "C:\Users...\ProteoWizard 3.0.23167.44089af 64-bit\msconvert.exe" --combineIonMobilitySpectra --filter "peakPicking vendor msLevel=1-" --filter "scanSumming precursorTol=0.05 scanTimeTol=5 ionMobilityTol=0.1 sumMs1=0" --filter "titleMaker ... File:"""^<SourcePath^>""", NativeID:"""^<Id^>""""

So given that it skips all the scans, and that it states that the precursor info is invalid, i was wondering what your settings were to generate the scan title, in other words what is your "titlemaker" part of your conversion command. I hope this makes sense. Also, please let me know if you have other suggestions for what could be wrong. :)

I have limited hands-on experience with timsTOF conversion to mzML, so I don't know how the titleMaker filter should be used. But I'd be surprised if that's the problem. I suspect something about the IM actually.

Can you share the mzML file here to have a look at?

Unfortunately I'm unable to share a file here. If you have an E-mail we could continue the conversation over we could maybe figure something out.

Alternatively I could try to compare the headers of your demo data with my data.

You can email me at [email protected].

Ok, the issue is that the scan titles in your mzML file are in the format merged=XX frame=XX scanStart=XXX scanEnd=XXX, whereas the DepthCharge parser expects a single scan number indicated by scan=XXX. The latter is ok when working with Thermo data, which is what we've mostly been doing so far. But of course not all scan titles are formatted that way, and PASEF is then an even slightly more special case.

The good news is that should be resolved by the pending DepthCharge upgrade (#350). Until that is fully integrated, I'll keep this issue open so that we can double-check that it gets fixed.

As a workaround for now, is it possible to modify the titleMaker filter? Alternatively, converting to MGF should also work, because for MGF we don't try to extract scan information from the spectrum title.

Hi @bittremieux

Thanks a lot for getting back to me and keeping me up to date. :)

I think the solution is in the headder, thus, for now i will leave it and await you update and or solution in #369 to work. If i may add one suggestion to the process though, it is this; please read the documentation for the "titleMaker" and its commands/syntax in msconvert (https://proteowizard.sourceforge.io/tools/msconvert.html), for you guys it could be advantageous to define a format that supports timsTOF and equiptment with those commands in mind. The result should be a command resembling what i showed above. The bonus here is that you can then add that example command to your readme wherefrom others can find the information as well.

Regarding mgf files, yes it's an option, and that worked when i tested it. However, from mgf files i am unable to estimates inteseties and thus abundances.

Thanks a lot for the tool and keep up the good work, i'll be keeping an eye on it. :)