Bayesian deconvolution of mass and ion mobility spectra: from binary interactions to polydisperse ensembles

Area of Science:

• Analytical Chemistry
• Biophysics
• Computational Biology

Background:

• Mass spectrometry interpretation is complex due to overlapping mass-to-charge (m/z) signals.
• Existing methods like peak assignment and fitting have limitations in handling complex spectra and computational demands.

Purpose of the Study:

• To develop a novel computational approach for deconvoluting mass spectra.
• To create user-friendly software (UniDec) for robust mass and charge separation.

Main Methods:

• A Bayesian deconvolution framework was employed to separate mass and charge dimensions.
• Incorporation of charge-state distribution into Bayesian prior probabilities.
• Development of UniDec software for automated spectral deconvolution.

Main Results:

• UniDec successfully deconvoluted mass spectra and ion mobility-mass spectra with minimal user intervention.
• The method effectively separated m/z spectra into physical mass and charge components.
• Demonstrated utility in analyzing lipid binding, subunit exchange dynamics, and polydispersity in protein assemblies and Nanodiscs.

Conclusions:

• The proposed Bayesian approach and UniDec software offer a rapid, robust, and flexible solution for mass spectral deconvolution.
• This method significantly facilitates the analysis of complex ion mobility and mass spectrometry data.
• Broad applicability across various biological systems, including membrane proteins and lipoprotein complexes.