The Era of Intelligent Drug Discovery: A Review of AI, ML, and DL in Computer-Aided Drug Design

Abstract

There is a lot of interest in this fledgling industry because of the potential for computer-aided drug design (CADD) to expedite and lower the cost of the drug development process. It normally takes 10 to 15 years for a drug to reach the market, and the research involved in its creation is time and money-consuming. CADD has had a significant impact on this area of study. By merging CADD with AI, ML, and DL technologies to manage massive amounts of biological data generated by combinatorial chemistry, the time and cost associated with the drug development process have also been reduced. These methods help in the identification of prospective drug candidates, target identification, and therapeutic molecule optimization by utilizing large-scale data analysis, predictive modeling, and pattern recognition. To find significant patterns and links, AI-driven algorithms can analyze enormous volumes of biological data, including genomes, proteomics, and chemical databases. Support vector machines, random forests, and neural networks are examples of ML. In DL, a subset of ML, deep neural networks are used to train a hierarchical data representation, which improves performance on tasks like protein-ligand binding affinity prediction, de novo drug creation, and virtual screening. Problems persist despite their achievements. In this review, we'll discuss the many phases of the drug development process and how these tactics could be used. Additionally, we will describe several in silico techniques that are employed and how they interact.