Discovery, characterization, and generative deep-learning-based humanization of therapeutic antibody candidates

In Part I of this webinar, we describe discovery of therapeutic, diagnostic, and tool antibody candidates against the Receptor for Advanced Glycation End products (RAGE) in mouse, rat, and rabbit hosts using the Beacon® Optofluidic instrument. We show how the Carterra LSA with HT-SPR technology was used to measure binding kinetics of the anti-RAGE antibodies — including multiple rabbit antibodies with sub 10 pM affinity — and to run epitope binning on candidates from each of the three host species in a single experiment. Finally, we conclude Part I by illustrating use of ENPICOM’s software platform for biologics discovery to select lead candidates by combining sequence diversity, wet-lab data, and in-silico generated developability profiles of all antibodies.

In Part II of this webinar, we introduce a next-generation, multi-species antibody humanization service powered by ENPICOM’s novel AIGX generative deep learning model. Through case-study data on anti-RAGE antibodies, we show that GenovacAI generates humanized antibody sequences at a much greater success rate — measured by percent germline identity and retention of binding affinity — than traditional tools, including CDR grafting and publicly available machine learning algorithms. The measurably improved functionality of GenovacAI effectively enables humanization of large numbers of candidate antibodies immediately post-discovery, resulting in reduced development costs and a shortened timeline to IND filing.