Artificial intelligence in therapeutic antibody design: Advances and future prospects

In the few years since AlphaFold 2 revolutionized protein structure prediction, AI technologies have demonstrated strong potential for practical application in therapeutic antibody development, a key area in the pharmaceutical industry. This mini-review provides a concise overview of AI-driven approaches designed to precisely optimize antibody properties critical for successful therapeutics. In particular, protein structure prediction-based antibody design AI is advancing rapidly, facilitating the effective targeting of protein hotspots, as demonstrated in a few reported cases. These advancements are expected to streamline experimental workflows, reduce reliance on trial-and-error screening, and enable the efficient discovery of novel molecules that would be challenging to identify through traditional methods. Additionally, this review explores emerging AI methodologies aimed at optimizing Fc function, immunogenicity, and developability, offering insights into future directions in the field.