A Protein Identification Algorithm Optimization for Mass Spectrometry Data using Deep Learning

Abstract

Protein sequence database search is one of the most commonly used methods for protein identification in shotgun proteomics. In tradition, searching a protein sequence database is usually required to construct the theoretical spectrum for each peptide at first, which only considers the information of mass-to-charge ratio at present. However, the information related to isotope peak intensity is neglected. Thanks to the rapid development of artificial intelligence technique in recent years, deep learning-based MS/MS spectrum prediction tools have showed a high accuracy and great potentials to improve the sensitivity and accuracy of protein sequence database searching. In this study, we used a deep learning model (pDeep2) to predict the theoretical mass spectrum of all peptides and applied it to a database searching tool (DeepNovo), thus improving the sensitivity and accuracy of peptide identification.