Abstract:Protein glycosylation plays a significant role in numerous physiological and pathological cellular functions. Glycoproteomics based on liquid chromatography-tandem mass spectrometry (LC-MS/MS) studies the protein glycosylation on a proteome-wide scale to get combinational information on glycosylation site, glycosylation level and glycan structure. However, the current sequence searching-based methods for glycoproteomics often fall short in glycan structure determination due to the limited occurrence of structure-determining ions. While spectral searching methods can utilize fragment intensity information to facilitate the identification of glycopeptides, its application is hindered by the difficulties in spectral library construction. In this work, we present DeepGP, a hybrid deep learning framework based on Transformer and graph neural network (GNN), for the prediction of MS/MS spectra and retention time of glycopeptides. Two GNN modules are utilized to capture the branched glycan structure and predict glycan ions intensity, respectively. Additionally, a pre-training strategy is implemented to alleviate the insufficiency of glycoproteomics data. Testing on multiple biological datasets, we demonstrate that DeepGP can predict MS/MS spectra and retention time of glycopeptides closely aligning with the experimental results. Comprehensive benchmarking of DeepGP on synthetic and biological datasets validates its effectiveness in distinguishing similar glycoforms. Remarkably, DeepGP can differentiate isomeric glycopeptides using MS/MS spectra without diagnostic ions. Based on various decoy methods, we demonstrated that DeepGP in combination with database searching can significantly increase the detection sensitivity of glycopeptides. We outlook that DeepGP can inspire extensive future work in glycoproteomics.

Deep learning prediction of electrospray ionization tandem mass spectra of chemically derived molecules

A Systematic Survey of Chemical Pre-trained Models

Phenotype Prediction using a Tensor Representation and Deep Learning from Data Independent Acquisition Mass Spectrometry

Pdeep: Predicting MS/MS Spectra of Peptides with Deep Learning.

Predicting A Molecular Fingerprint From An Electron Ionization Mass Spectrum With Deep Neural Networks

DMSS: an Attention-Based Deep Learning Model for High-Quality Mass Spectrometry Prediction

Deep peak property learning for efficient chiral molecules ECD spectra prediction

Advancing the Prediction of MS/MS Spectra Using Machine Learning

Emergence of molecular structures from repository-scale self-supervised learning on tandem mass spectra

Deep Learning Prediction of Glycopeptide Tandem Mass Spectra Powers Glycoproteomics

A Decade of Computational Mass Spectrometry from Reference Spectra to Deep Learning

Prediction of Electron Ionization Mass Spectra Based on Graph Convolutional Networks

A deep learning model for predicting selected organic molecular spectra.

Large-Scale Prediction of Collision Cross-Section with Very Deep Graph Convolutional Network for Small Molecule Identification

Chemical Space Localization for Unknown Metabolite Annotation via Semantic Similarity of Mass Spectral Language

Deep learning based CETSA feature prediction cross multiple cell lines with latent space representation

MS/MS Spectrum Prediction for Modified Peptides Using Pdeep2 Trained by Transfer Learning

DeepCSNet: a deep learning method for predicting electron-impact doubly differential ionization cross sections

Test-Time Training for Deep MS/MS Spectrum Prediction Improves Peptide Identification.

Spectroscopy-Guided Deep Learning Predicts Solid-Liquid Surface Adsorbate Properties in Unseen Solvents.