Integrating Protein-Protein Interaction Networks and Machine Learning for Drug Target Identification

In this study, we combined the STRING dataset of mouse protein-protein interactions (PPIs) with a customdeveloped algorithm designed for simulation and analysis. By utilizing this approach, we conducted an in-depth exploration of the genes within the PPI network and identified key biological processes and signaling pathways critical for cellular mechanisms. To assess the involvement of these genes in biological processes, molecular functions, and related pathways, we applied Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations. The analysis identified important genes, pathways, and functions, providing valuable insights for disease research and drug target identification. Furthermore, the algorithm incorporated machine learning models, particularly Random Forest (RF) for PPI prediction, and Frequent Pattern Growth (FP-Growth) for discovering frequent gene-drug associations. These advanced computational methods proved effective in uncovering critical interactions, enhancing the efficiency of drug discovery, and offering a deeper understanding of potential drug targets and their biological relevance.