Traditional drug discovery processes are typically expensive, time-consuming, and have a very high failure rate. Artificial intelligence (AI) is currently reshaping this field in unprecedented ways, promising to significantly improve the efficiency and success rate of drug development. This article systematically compares and analyzes the application of AI for two major drug types: small molecule vs. peptide drugs. It explores their applications in several key stages of drug development, including virtual screening, lead compound optimization, de novo drug design, ADMET (absorption, distribution, metabolism, excretion, and toxicity) property prediction, and chemical synthesis planning. While both drug types benefit from AI-driven approaches, fundamental differences exist in molecular representation, data availability, key challenges, and model adaptability. For small molecule drugs, AI focuses on drug efficacy, synthetic feasibility, and accurate structure–activity relationship prediction. In contrast, for peptide drugs, AI faces more unique biological challenges, such as inherent flexibility, complex biological functions, stability, and immunogenicity. Finally, this article provides a forward-looking perspective on the future of AI-driven drug discovery, highlighting the immense potential of basic models, multimodal integrated systems, and autonomous discovery platforms, which will collectively drive the next wave of precision drug development.
Lin et al. (Mon,) studied this question.