A Systematic Literature Review of the Current Status and Future Prospects of Machine Learning Methods and Techniques Applied to Novel Drug Discovery

Abstract

Drug development is the hardest phase for the pharmaceutical industry because it is extremely costly and time consuming. Though, due to the growing demand to produce safe and innovative medicines faster and more cost-effectively, the scientific community changed its objective into enhancing the lead identification and the lead optimization at the early discovery phase. This could be achieved using recent intelligent technologies that allow virtual screening as well as quantitative structure-activity relationship (QSAR) modeling to define the possible relationships between chemical compounds and biological activities. Among recent technologies, artificial intelligence (AI) has been introduced as a powerful solution to address problems related to drug discovery and development. In particular, machine learning (ML) has been meaningfully instrumental in the production of new drug candidates. In this work, we review the fundamental principles of machine learning algorithms, study and discuss their application and current issues in drug development.