FR: Ion mobility support?

[rickhelmus]

Hello,

While looking for options to load mzML files from C/C++ I stumbled upon MSToolKit :-)

There seems some ion mobility support in mzParser, but his seems unused at the moment in MSToolKit. I was wondering if adding support to load IMS data is something on the radar?

Thanks,
Rick

[mhoopmann]

Hi Rick, I can certainly help build out that support. Do you have some example mzML data to share and an idea of what you need parsed out from it? Cheers, Mike

[rickhelmus]

Hi Mike,

Thanks for your quick reply, and great to hear you want to help out.

Currently, I'm mainly interested in processing timsTOF data. Here a typical file seems to add a third binary data array (MS:1003006), which represents the mobility for each mass peak. Normally the files are huge, but I made a small subset here:
20220923_TIMS03_PaSk_SA_4prot_200fmol_Evo05_3min_IM0713_classical_SyS_4MS_wCE_S4-G3_1_32594.mzML.gz (the original file comes from here, which I converted with msConvert).

Later on I would also be able to handle IMS data from other instruments. I don't have much experience with that yet, but looking at an Agilent file it seems each spectrum is tagged with a mobility (MS:1002476) instead of adding an extra dimension. An example file can be found here: https://massive.ucsd.edu/ProteoSAFe/dataset_files.jsp?task=749e436db868410383159b450b470eff#%7B%22table_sort_history%22%3A%22main.collection_asc%22%7D

It would be great if it would be possible to load the raw MS spectra with mobility data from both formats.

Thanks,
Rick

[mhoopmann]

Hi Rick,
I put in a small update to get you started. It should work for TimsTOF data. The Spectrum object contains a new function Spectrum::getInverseReducedIonMobility() which should return a non-zero value when the spectrum contains ion mobility data. To get at the ion mobility values, use Spectrum::atIM(size_t index), where the dimensions of the ion mobility array fall under Spectrum::size().

I'll probably need more time to extend the new functionality to non-TimsTOF data. Also, I haven't implemented the ion mobility functionality to mzMLb data yet, but that is also on the list. I'll post another update when these features are finished.
Cheers,
Mike

[rickhelmus]

Hi Mike,

This is really great, many thanks! All seems to work well with a small toy example. Looking forward to integrate this further later :-) I'm wondering now what the values mean what getInverseReducedIonMobility() return, but this is just a curiousity and maybe only Bruker can answer that.

I hadn't even thought about mzMLb support, but now I'm quite curious how that would perform.

Anyway, thanks again, and looking forward to your next update!