DeepSiPPI: Enhancing Protein-Protein Interaction Prediction through Siamese Neural Networks and Sequence-to-Image Transformation

This study focuses on the analysis and prediction of Protein-Protein Interactions (PPI) withinbiological systems. We propose an innovative methodology employing a Siamese Neural Networkarchitecture with a Convolutional Neural Network (CNN) as its underlying framework. Notably, weintroduce a novel preprocessing step involving the conversion of protein sequences into images throughthe application of a robust statistical method. Subsequently, these transformed representations areutilized as input data for the Siamese Neural Network, a choice motivated by its intrinsic capacity foreffective feature extraction. The discerning ability of the Siamese Neural Network proves instrumentalin discerning subtle patterns and features crucial for the identification of protein interactions. Thepresented approach not only showcases its utility in refining interaction predictions but also underscoresthe potential for advancing the comprehension of intricate protein networks, thereby contributing to thebroader landscape of bioinformatics research .

کلیدواژه ها:

Protein-Protein Interaction ، Siamese Neural Network ، Protein Representation ، Sequenceto-Image Transformation ، Bioinformatics