Artificial intelligence in rational drug discovery: optimization of lead compounds and in-silico prediction of molecular interactions for faster pharmaceutical development.

The conventional drug discovery process is intricate, expensive and time-consuming with high attrition rates. Addressing these inefficiencies has spelled the dawn of artificial intelligence (AI) and machine learning (ML) as game-changers in rational drug design. The review highlights the revolutionary potential of AI techniques such as deep learning and generative models in drug discovery and development. We discuss how AI improves protein-ligand interaction predictions and quantitative structure-activity relationship (QSAR) modelling, as well as virtual screening, in drug discovery. In addition, AI can significantly accelerate lead optimization by predicting critical absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties early in the process, allowing for prioritization of promising candidates before time-consuming experimental testing. Although these advancements have been made there still remain several limitations to how broad AI systems are applied in the real world, ranging from data bias and model reliability to the non-interpretability of "black box" neural networks. To overcome these regulatory and ethical challenges, Explainable AI (XAI) must be implemented. In the end, the combination of AI with multi-omics and other technologies such as quantum computing will bring personalized therapy at a much more specific level and significantly shorten the development and research times and costs.