Expansion of open-source tools handling IMS data.

[swillems]

Abstract

Data Independent Acquisition (DIA) creates a “permanent record of everything” that is ideal for reanalysis, arguably the primary goal of public data repositories.

Within ProteomeXchange, Waters' Synapt is among the top 10 most used instruments. This instrument allows for a DIA technique called HDMSE. Herein, ions are separated in seconds by LC, in milliseconds by Ion Mobility Separation (IMS) and in nanoseconds by MS. Thus, detected ions have four dimensions: retention time, drift time, m/z and intensity. Each MS scan is followed by a single MSMS scan without any precursor selection and fragments can be assigned to precursors based on retention and drift time profiles.

Unfortunately, few software packages exist that can handle this data format. Examples are the open-source packages ISOQuant and SynapteR and the commercial Progenesis QIP software, but all three rely on data preprocessing with Waters' commercial PLGS software. While ProteoWizard can convert raw Waters data to mzML, most (open-source) tools such as OpenMS and TPP are currently unable of fully handling the IMS dimension in mzML, let alone aligning fragments to precursors. In conclusion, even basic tools are lacking to start building an open-source workflow that allows (re)analysis of Waters' HDMSE data.

Technical details

• Project outcome
  • Create a Peak Picker that converts an IMS-enhanced profile mzML to an IMS-enhanced centroided mzML
  • Create a Feature Finder that deisotopes an IMS-enhanced centroided mzML and finds features.
• Required familiarity
  • Python
  • mzML format
• Useful familiarity
  • ProteoWizard
  • OpenMS or TPP
  • PLGS
  • ISOQuant or SynapteR or Progenesis QIP
  • ProteomeXchange
  • HDMSE and/or Waters raw format
  • Peptide3D and/or Apex3D algorithms
• Dataset
  • PXD001240 for final testing
  • A smaller technical dataset for developmental purposes can be provided by the Laboratory of Pharmaceutical Biotechnology

Contact information

Sander Willems
Laboratory of Pharmaceutical Biotechnology, Ghent University, Ghent, 9000, Belgium
[email protected]

[timosachsenberg]

Sounds interesting, Are mzMLs with profile-spectra available?

[swillems]

Hi Timo. Thanks for the support!
As there seem to be no follow-up tools for IMS-enhanced mzML files, I haven't made much effort to find them in public data repositories yet. I'm guessing they a rare in proteomics, but it is possible they do exist for metabolomics. However, there are plenty of raw datasets available (both publicly and in-house from our own Synapt G2-SI) and I have successfully been able to convert some of them with ProteoWizard, so there shouldn't be any problem to get obtain profile mzMLs.