Abstract:Protein-protein interaction plays an important role in all biological systems. The binding affinity between two protein binding partners reflects the strength of their association, which is crucial to the elucidation of the biological functions of these proteins and also to the design of protein-based therapeutic agents. In recent years, many studies have been conducted in an effort to improve the ability to predict the binding affinity of a protein-protein complex. Different sequence and structural features have been adopted in the prediction, but the surface or interface areas of the protein-protein complex were often not given adequate consideration. In the present study, different types of interface and surface areas in the protein-protein complex were used to construct or train linear, nonlinear or mixed models using linear regression and artificial neural network to predict the binding affinity of protein-protein interactions. The relative importance of the different types of areas in the selected models for affinity prediction was analyzed using variable-controlling approach. In terms of performance, the best area-based binding affinity predictors appeared to be superior or at least comparable to the widely-used predictors PRODIGY (a contacts-based predictor) and LISA (Local Interaction Signal Analysis). This work highlights the importance of interface and surface areas in protein-protein binding interactions. It also sheds light on the more suitable computational approaches that may aid in solving some of the scientific and technical issues associated with protein-protein binding affinity prediction. SIGNIFICANCE: Protein-protein interactions are ubiquitous in living systems. Protein-protein binding affinity is a metric that estimates the binding strength between two protein binding partners. Reliable information on their binding affinity is of great value in understanding complex biological processes as well as in designing protein-based therapeutics. In this work, the interface and surface areas in protein-protein interaction are explored with respect to their relative importance in better predicting the protein-protein binding affinity. The results from this study showed that different types of areas contribute importantly to protein-protein interactions and thus should be jointly considered in an explicit manner to improve affinity predictions. In addition, the effective application of interface and surface areas may also facilitate the simulation of the protein folding and binding processes.

Protein-Protein interaction affinity prediction based on interface descriptors and machine learning

Deep Learning Frameworks for Protein–protein Interaction Prediction

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

Importance of interface and surface areas in protein-protein binding affinity prediction: A machine learning analysis based on linear regression and artificial neural network

Protein-Protein Interaction Interface Residue Pair Prediction Based on Deep Learning Architecture

Accurate Prediction of Protein-Protein Interactions by Integrating Potential Evolutionary Information Embedded in PSSM Profile and Discriminative Vector Machine Classifier

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

Highly Accurate Prediction of Protein-Protein Interactions Via Incorporating Evolutionary Information and Physicochemical Characteristics.

Advances in Computational Methods for Protein–Protein Interaction Prediction

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

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

Predicting Protein Interactions Using a Deep Learning Method-Stacked Sparse Autoencoder Combined with a Probabilistic Classification Vector Machine.

Prediction of transient and permanent protein interactions using AI methods

DeepPPI: Boosting Prediction of Protein-Protein Interactions with Deep Neural Networks.

Prediction of Protein-Protein Interactions from Amino Acid Sequences with Ensemble Extreme Learning Machines and Principal Component Analysis

Protein-protein interaction prediction via structure-based deep learning

Prediction of Protein–Protein Interactions Based on Integrating Deep Learning and Feature Fusion

Protein-Protein Binding Affinity Prediction Based On Wavelet Package Transform And Two-Layer Support Vector Machines

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

Improving protein-protein interaction site prediction using deep residual neural network

Classification and prediction of protein–protein interaction interface using machine learning algorithm