Sanofi is advancing its AI-based protein design program from proof-of-concept to real-world drug discovery. Norbert Furtmann, PhD, head of biologics AI and design at Sanofi, outlines a three-pronged approach: foundational models, property prediction and virtual screening, and de novo design.
The foundational models team creates sequence- and structure-based models trained on public and in-house data. The property prediction team develops models to predict essential properties such as stability and solubility, running large in silico screens on hundreds of thousands of sequences. This process is crucial for multispecifics, where extensive lab testing is impractical due to the vast combinatorial design space.
Sanofi’s de novo design team uses AI to create antibodies or nanobodies against target antigens. Furtmann notes that successful proof-of-concept studies are leading to program scale-up. The integration of AI with lab workflows aims to optimize the design process and improve the development of complex multispecific molecules.
As companies increasingly adopt AI in drug design, those with advanced data integration capabilities may gain a competitive edge in developing innovative therapeutics.
Source: genengnews.com · Image: AI golem waiting for tasks and providing advice by Prototyperspective, CC0, via Wikimedia Commons
Get started today with Solo access →
