**Abstract**

The drug discovery process from hit-to-lead has been a challenging task that requires simultaneously optimizing numerous factors from maximizing compound activity, efficacy to minimizing toxicity and adverse reactions. Recently, the advance of artificial intelligence technique enables drugs to be efficiently purposed *in silico* prior to chemical synthesis and experimental evaluation. In this chapter, we present fundamental concepts of artificial intelligence and their application in drug design and discovery. The emphasis will be on machine learning and deep learning, which demonstrated extensive utility in many branches of computer-aided drug discovery including de novo drug design, QSAR (Quantitative Structure–Activity Relationship) analysis, drug repurposing and chemical space visualization. We will demonstrate how artificial intelligence techniques can be leveraged for developing chemoinformatics pipelines and presented with real-world case studies and practical applications in drug design and discovery. Finally, we will discuss limitations and future direction to guide this rapidly evolving field.

**Keywords:** artificial intelligence, chemoinformatics, data mining, drug discovery

## **1. Introduction**

The path of drug discovery from small molecule ligands to drugs that can be utilized clinically has been a long and arduous process. Starting with a hit compound, the drugs need to be evaluated through multiple *in vitro* and cell-based assays to improve the mechanism of actions followed by mouse models to demonstrate appropriate *in vivo* and transport properties. Mechanistically, the drugs not only need to exert enough binding affinity to the disease targets, but also necessitate proper transport through multiple physiological barriers to enable access to these targets. Other problems like chemical toxicity, often induced by off-targets interactions with unintended proteins as well as pharmacogenetic, where genetic variation influences drug responses all need to be considered in drug design. Therefore, these multifaceted problems in drug discovery often posed significant challenges for drug designers. Recently, the rise of artificial intelligence approach saw potential solutions to these challenges. A sub-umbrella of artificial intelligence called machine-learning has taken a central stage in many R&D sectors of pharmaceutical companies that allows drugs to be developed more efficiently and at the same time mitigate the cost associated with the required experiments [1]. Given some observations of chemical data, machine learning can be used to construct a predictor by learning compound properties from extracted features of compound structures and interactions. Because this approach does not require a mechanistic

understanding of how drugs behave, many compound properties like binding affinity and other transport and toxicity problems can be accurately forecasted in this way before they are synthesized [2]. Furthermore, by simultaneously tackling the Pharmacokinetics/Pharmacodynamics (PK/PD) problems using artificial intelligence, we can expect that the effort and time required to bring a drug from bench to bedside can be substantially reduced. In this regard, the artificial intelligence approach has now become an essential tool to facilitate the drug discovery process.
