In Silico Top-Down Fourier Transform Mass Spectra to Advance Proteoform Analysis Methods

Abstract

Top-down (TD) Fourier transform mass spectrometry (FTMS) of proteins generates highly information-rich mass spectra. However, the resulting spectral complexity can hinder data interpretation and method applicability. Here, we apply transient-mediated, instrument-specific simulations of protein TD mass spectra with user-defined product ion identities. The workflow integrates sequence-based product ion generation, empirically informed fragmentation statistics, charge location modeling, and user-specified FT processing approach. The simulations capture key FTMS characteristics, including m/z-dependent resolution, peak interference, and realistic peak shapes. As a proof of concept, we simulated carbonic anhydrase II TDMS data sets under conditions reported for corresponding 21 T ion cyclotron resonance (ICR) FTMS electron transfer dissociation experiments. The simulated mass spectra reproduced key experimental descriptors of spectral complexity, including peak density, spectral dynamic range, and isotopic structures. Using data sets with increasing complexity, we performed an initial evaluation of TDMS annotation performance and observed an expected decrease in product ion annotation rates from c-ion-only mass spectra to the most complex data sets containing a-, c-, y-, and z-type product ions, reflecting the impact of spectral congestion on confident assignment. These results demonstrate that realistic in silico TDMS data sets can support systematic investigation of spectral properties in TDMS and facilitate the development and evaluation of proteoform-specific data analysis approaches. The initial data set of simulated mass spectra and transients is available through ProteomeXchange under identifier PXD075270.