Development of a Python-based electron ionization mass spectrometry amino acid and peptide fragment prediction model

TL;DR

This paper presents a Python-based model to predict peptide fragment patterns under harsh electron ionization mass spectrometry conditions.

Contribution

A novel Python-based model for predicting peptide fragments under electron ionization mass spectrometry is developed and experimentally validated.

Findings

The model accurately predicts fragments for well-defined amino acids.

Testing revealed cases requiring additional mechanistic information for improved predictions.

Abstract

The increased fragmentation caused by harsher ionization methods used during mass spectrometry such as electron ionization can make interpreting the mass spectra of peptides difficult. Therefore, the development of tools to aid in this spectral analysis is important in utilizing these harsher ionization methods to study peptides, as these tools may be more accessible to some researchers. We have compiled fragmentation mechanisms described in the literature, confirmed them experimentally, and used them to create a Python-based fragment prediction model for peptides analyzed under direct exposure probe electron ionization mass spectrometry. This initial model has been tested using single amino acids as well as targeted libraries of short peptides. It was found that the model does well in predicting fragments of peptides composed of amino acids for which the model is well-defined, but several cases where additional mechanistic information needs to be incorporated have been identified.

Figures