DCSE:Double-Channel-Siamese-Ensemble model for protein protein interaction prediction.

Area of Science:

• Computational biology
• Bioinformatics
• Genomics

Background:

• Protein-protein interactions (PPI) are crucial for biochemical processes.
• Accurate PPI prediction aids in understanding protein functions.
• Existing PPI prediction methods are often time-consuming and lack precision.

Purpose of the Study:

• To develop an efficient and accurate computational model for predicting protein-protein interactions.
• To introduce a novel sequence-based approach for PPI prediction.

Main Methods:

• Developed DCSE (Double-Channel-Siamese-Ensemble), a novel sequence-based computational approach.
• Utilized a natural language processing-based encoding layer, treating amino acids as words mapped to vectors.
• Employed Multilayer Convolutional Neural Network (MCN) and Multilayer Bidirectional Gated Recurrent Unit with Convolutional Neural Networks (MBC) for feature extraction.
• Implemented siamese and ensemble network structures within MCN and MBC to enhance model performance.

Main Results:

• DCSE demonstrated superior performance compared to four machine learning and three deep learning models across all evaluation metrics.
• Achieved high Accuracy (0.9303), Precision (0.9091), Recall (0.9452), and MCC (0.8609).
• Showcased excellent performance on imbalanced datasets and successful transferability to other species.

Conclusions:

• The DCSE model significantly outperforms existing methods in PPI prediction.
• NLP-based coding and siamese/ensemble network structures effectively enhance feature extraction and prediction accuracy.
• The developed model offers a robust and efficient solution for predicting protein-protein interactions.