De novo design is not a magic bullet, but it's a powerful new tool. Major pharmaceutical companies report it successfully generates binders for difficult targets where conventional methods like immunization have failed, effectively closing critical gaps in the discovery pipeline.
The primary barrier to implementing AI for antibody developability isn't the tech, which has been available for over a decade. MIT's Bernhard Trout states the real failure point is a lack of sustained corporate commitment, as key personnel are frequently reassigned to other projects, causing initiatives to stall.
Instead of just augmenting existing wet lab workflows with AI, Sanofi's Norbert Furtman advocates for a paradigm shift. He suggests R&D leaders should design future workflows to be AI-driven from the start, with a customized wet lab built to serve as the 'perfect counterpart' to the in-silico tools.
AI is delivering tangible results now. An internal Eli Lilly study showed that using an AI-enabled triaging workflow for developability and structural diversity in early discovery has significantly reduced the number of 'surprises' and liabilities for molecules entering later development stages.
The lack of comparable developability data is a major bottleneck. Natural Antibody's CEO suggests a 'walk before you can run' approach: instead of accounting for all variables, the industry should create a foundational dataset under a single condition. This focused dataset has proven transferable predictive power.
While AI models are effective for developability properties like stability, they fall short on predicting function. Sanofi's Norbert Furtman notes that generalized affinity prediction is a 'holy grail' problem, and predicting interference with a biological pathway is even harder, as function is not solely explained by structure.
De novo AI is proving its value against notoriously difficult targets. Panelists from major pharmaceutical companies confirmed that these methods are achieving early, promising successes against targets like GPCRs, which have historically been challenging for conventional antibody discovery platforms.
A simple but critical data gap is hampering AI models. Most labs measure antibody affinity at room temperature for convenience. However, Andrew Buchanan argues this is not translationally relevant. To build effective predictive models, data must be generated at 37°C, the temperature where the drug will actually function.