Artificial intelligence in antibody-drug conjugate development

Abstract

Antibody-drug conjugates (ADCs) offer a promising approach for targeted cancer treatment. Progress, however, is constrained by the combinatorial complexity of design, toxicity and side effects, and variable clinical benefit across indications. Effective ADCs require rational matching of the antibody, linker, and payload to achieve stability in circulation and tumor-specific release, which makes development time- and cost-intensive. Artificial intelligence (AI) is shifting ADC development from empirical trial-and-error to data-driven, closed-loop engineering. By integrating sequence (for antibodies), structural, and molecular dynamics (MD) features of ADC components, AI models can accelerate target selection, conjugate optimization, and patient-response prediction. This review synthesizes advances in AI-driven ADC development across preclinical and clinical phases, highlights representative case studies and industry platforms, and outlines opportunities for AI-enabled next-generation ADCs.

Keywords: antibody–drug conjugates; artificial intelligence; drug discovery; machine learning.