In Silico Approaches to Deliver Better Antibodies by Design: The Past, the Present and the Future

Abstract: The recognition of the importance of drug-like properties beyond potency to reduce clinical attrition of biologics has driven significant progress in the development of in vitro and in silico tools for developability assessment of antibody sequences. It is now routine to identify and eliminate or optimize antibody hits with poor developability profiles. To further accelerate discovery timelines and reduce clinical and non-clinical development attrition rates, more proactive in silico approaches to design sequence spaces with favorable developability profiles are required. From pragmatically front-loading structure based drug design for developability, to combining next generation sequencing with machine learning to shape screening libraries, to adapting the use of artificial intelligence and deep learning for immunoglobulins, we review herein progressively more proactive approaches to developability by design.