Abstract:The study of protein-protein interactions (PPIs) is important in understanding the function of proteins. However, it is still a challenge to investigate the transient protein-protein interaction by experiments. Hence, the computational prediction for protein-protein interactions draws growing attention. Statistics-based features have been widely used in the studies of protein structure prediction and protein folding. Due to the scarcity of experimental data of PPI, it is difficult to construct a conventional statistical feature for PPI prediction, and the application of statistics-based features is very limited in this field. In this paper, we explored the application of frustration, a statistical potential, in PPI prediction. By comparing the energetic contribution of the extra stabilization energy from a given residue pair in the native protein with the statistics of the energies, we obtained the residue pair's frustration index By calculating the number of residue pairs with a high frustration index, the highly frustrated density, a residue-frustration-based feature, was then obtained to describe the tendency of residues to be involved in PPI. Highly frustrated density, as well as structure-based features, were then used to describe protein residues and combined with the long short-term memory (LSTM) neural network to predict PPI residue pairs. Our model correctly predicted 75% dimers when only the top 2 parts per thousand residue pairs were selected in each dimer. Our model, which considers the statistics-based features, is significantly different from the models based on the chemical features of residues. We found that frustration can effectively describe the tendency of residue to be involved in PPI. Frustration-based features can replace chemical features to combine with machine learning and realize the better performance of PPI prediction. It reveals the great potential of statistical potential such as frustration in PPI prediction.

PF2PI: Protein Function Prediction Based on AlphaFold2 Information and Protein-Protein Interaction

DeepTrio: a Ternary Prediction System for Protein-Protein Interaction Using Mask Multiple Parallel Convolutional Neural Networks

A Deep Learning Framework for Improving Protein Interaction Prediction Using Sequence Properties

Using Two-dimensional Principal Component Analysis and Rotation Forest for Prediction of Protein-Protein Interactions

Enhanced Protein-Protein Interaction Discovery via AlphaFold-Multimer

Enhanced prediction of protein functional identity through the integration of sequence and structural features

DeepSG2PPI: A Protein-Protein Interaction Prediction Method Based on Deep Learning

Protein Function Prediction by Collective Classification with Explicit and Implicit Edges in Protein-Protein Interaction Networks

Residue-Frustration-Based Prediction of Protein-Protein Interactions Using Machine Learning

Improved prediction of protein-protein interactions using AlphaFold2

Prediction Of Protein-Protein Interactions Using Subcellular And Functional Localizations

Advancing the prediction accuracy of protein-protein interactions by utilizing evolutionary information from position-specific scoring matrix and ensemble classifier

Enhancing Protein Function Prediction Performance by Utilizing AlphaFold-Predicted Protein Structures

A New Protein-Protein Interaction Prediction Algorithm Based on Conditional Random Field.

Struct2GO: Protein Function Prediction Based on Graph Pooling Algorithm and AlphaFold2 Structure Information.

An Integration of Deep Learning with Feature Embedding for Protein–protein Interaction Prediction

Improving protein function prediction using domain and protein complexes in PPI networks

Application of the AlphaFold2 Protein Prediction Algorithm Based on Artificial Intelligence

Prot2GO: Predicting GO Annotations from Protein Sequences and Interactions.

Enhanced Protein Fold Prediction Method Through a Novel Feature Extraction Technique.

Protein-Protein Interaction Prediction Model Based on ProtBert-BiGRU-Attention