Journal of Biomedical Science, Год журнала: 2025, Номер 32(1)
Опубликована: Май 9, 2025
Abstract The integration of high-throughput experimentation and machine learning is transforming data-driven antibody engineering, revolutionizing the discovery optimization therapeutics. These approaches employ extensive datasets comprising sequences, structures, functional properties to train predictive models that enable rational design. This review highlights significant advancements in data acquisition feature extraction, emphasizing necessity capturing both sequence structural information. We illustrate how models, including protein language are used not only enhance affinity but also optimize other crucial therapeutic properties, such as specificity, stability, viscosity, manufacturability. Furthermore, we provide practical examples case studies demonstrate synergy between experimental computational accelerates engineering. Finally, this discusses remaining challenges fully realizing potential artificial intelligence (AI)-powered pipelines expedite development.
Язык: Английский