Advancements in Drug Discovery Through Machine and Deep Learning Concepts in Early-Stage of Drug Discovery for Apoptosis

Within the rapidly shifting field of drug discovery, ML and DL techniques have opened new directions for identifying potential therapeutic compounds with higher efficiency and accuracy. The present work explores the integration of ML and DL algorithms for drug discovery in apoptosis, by utilizing PaDEL and Lipinski descriptors for feature extraction. PaDEL is widely used open-source software that helps in generating the molecular descriptors and fingerprints of a chemical compound while Lipinski's descriptors depend on the “Rule of Five” which explores critical analysis of drug-likeness of molecules. The present article provides a comparative analysis of different ML and DL algorithms which include SVM, RF, GBM, KNN, ANN, and GNN. The performance of each one is evaluated based on key metrics such as adjustment of r2 score and mean square error. This presented work shall help in predicting the chemical compounds at the early stage of drug discovery using the QSAR method. Computed results are depicted graphically.