Abstract:Peptides are a powerful class of molecules that can be applied to a range of problems including biomaterials development and drug design. Currently, machine learning-based property prediction models for peptides primarily rely on amino acid sequence, resulting in two key limitations: first, they are not compatible with non-natural peptide features like modified sidechains or staples, and second, they use human-crafted features to describe the relationships between different amino acids, which reduces the model’s flexibility and generalizability. To address these challenges, we have developed PepMNet, a deep learning model that integrates atom-level and amino acid-level information through a hierarchical graph approach. The model first learns from an atom-level graph and then generates amino acid representations based on the atomic information captured in the first stage. These amino acid representations are then combined using graph convolutions on an amino acid-level graph to produce a molecular-level representation, which is then passed to a fully connected neural network for property prediction. We evaluated this architecture by predicting two peptide properties: chromatographic retention time (RT) as a regression task and antimicrobial peptide (AMP) as a classification task. For the regression task, PepMNet achieved an average R² of 0.980 across eight datasets, which spanned different dataset sizes and three liquid chromatography (LC) methods. For the classification task, we developed an ensemble of five models to reduce overfitting and ensure robust classification performance, achieving an area under the receiver operating curve (AUC-ROC) of 0.978 and an average precision of 0.981. Overall, our model illustrates the potential for hierarchical deep learning models to learn peptide properties without relying on human engineering amino acid features.

PepFormer: End-to-End Transformer-Based Siamese Network to Predict and Enhance Peptide Detectability Based on Sequence Only

PepNet: an interpretable neural network for anti-inflammatory and antimicrobial peptides prediction using a pre-trained protein language model

Pep-CNN: an Improved Convolutional Neural Network for Predicting Therapeutic Peptides

Predicting Protein-Peptide Binding Residues Via Interpretable Deep Learning.

PepMNet: A Hybrid Deep Learning Model for Predicting Peptide Properties Using Hierarchical Graph Representations

PepCA: Unveiling protein-peptide interaction sites with a multi-input neural network model

ConPep: Prediction of peptide contact maps with pre-trained biological language model and multi-view feature extracting strategy

PepGB: Facilitating peptide drug discovery via graph neural networks

PepLand: a large-scale pre-trained peptide representation model for a comprehensive landscape of both canonical and non-canonical amino acids

PepCNN deep learning tool for predicting peptide binding residues in proteins using sequence, structural, and language model features

A deep attention model for wide-genome protein-peptide binding affinity prediction at a sequence level

PepHarmony: A Multi-View Contrastive Learning Framework for Integrated Sequence and Structure-Based Peptide Encoding

Pep2Vec: An Interpretable Model for Peptide-MHC Presentation Prediction and Contaminant Identification in Ligandome Datasets

SiameseCPP: a sequence-based Siamese network to predict cell-penetrating peptides by contrastive learning

TP-LMMSG: a Peptide Prediction Graph Neural Network Incorporating Flexible Amino Acid Property Representation.

Predicting multifunctional peptides based on a multi-scale ResNet model combined with channel attention mechanisms

Protein-peptide binding residue prediction based on protein language models and cross-attention mechanism

PepExplainer: an explainable deep learning model for selection-based macrocyclic peptide bioactivity prediction and optimization

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

Enhancing Drug Peptide Sequence Prediction Using Multi-view Feature Fusion Learning

AttnPep: A Self-Attention-Based Deep Learning Method for Peptide Identification in Shotgun Proteomics.