Key points are not available for this paper at this time.
Artificial intelligence (AI) is a computer system that performs tasks that require learning, problem-solving and decision-making skills. It mimics humans' cognitive functions, which include pattern recognition. Vast amounts of data serve as "experience" for AI to train the rules and patterns from the fed datasets, just as humans learn from past knowledge. Then, AI analyses data using algorithms, and finally, it integrates the best output based on logic and probability. The integration of AI into drug discovery, on the other hand, is a revolutionary innovation as it supports and increases the efficacy of multiple stages in the traditional "trial and error" method, such as target identification, hit identification, lead optimisation, cell line model predictions, and even clinical development. Hence, incorporating AI into the pharmaceutical industry offers higher speed, reduced costs, and improved efficiency. However, AI-driven drug discovery is data-intensive; it relies on massive biological datasets, omics data, and chemical libraries. Therefore, noisy or biased data may mislead AI models, resulting in unreliable predictions in the lab. Although the application of AI in drug discovery accelerates the process, the integration of AI with human expertise must be carefully balanced. Rather than providing an exhaustive review of all existing AI applications in drug discovery, this commentary aims to present a perspective on the current opportunities, ongoing challenges, and future outlook of AI integration in pharmaceutical research, driving meaningful innovation in the coming fast-paced generation ahead.
Swee et al. (Fri,) studied this question.