Recursion and Genentech Advance First AI-Mapped Neuroscience Target

Recursion Pharmaceuticals and Genentech have entered a co-development program for a neuroscience small molecule after validating the first target discovered through a joint AI-driven "phenomap." The milestone marks a successful proof of concept for using massive-scale biological data to identify novel targets in the central nervous system.

Genentech exercised a validated target option to trigger the new discovery program, resulting in a $3 million milestone payment to Recursion. This brings the total payments from Roche and Genentech to Recursion to $216 million since the partnership began in 2021. The discovery resulted from a whole-genome CRISPR knockout map generated from more than one trillion lab-produced neuronal cells derived from human induced pluripotent stem cells (iPSCs). To identify the target, the team analyzed over 46 million cellular images across hundreds of features, utilizing AI foundational models and the BioHive-2 supercomputer. The transition from the start of target validation to the final validation package took 15 months.

The Challenge of Neuro-Discovery

Neuroscience has historically been one of the most difficult areas for drug discovery. The field has struggled to find novel targets beyond a few well-known pathways, leading to high failure rates. According to a 2006-2015 study cited by GEN, only 8.4% of neurology drug candidates that enter Phase I studies eventually reach approval.

To overcome these hurdles, Recursion employs a "phenomics" approach. This method combines a "Data Factory" for massive biological data generation with AI to identify unbiased gene-compound and gene-gene interactions within living human neurons. "Finding new targets in neuroscience has historically been challenging, and this milestone highlights our ability to uncover novel biology in areas where conventional approaches have struggled," said Najat Khan, PhD, CEO of Recursion.

Implications for the Industry

This milestone demonstrates a scalable method for expanding the universe of treatable neurological conditions. By moving from a trillion-cell map to a validated target, the partnership proves that AI-generated biological maps can successfully uncover viable targets in the CNS space. Christopher Winrow, PhD, Recursion VP of Neuroscience, noted that the ability to look at whole genome-wide knockouts, rather than just a few pathways of interest, is a "big game changer."

Future Outlook

This program is part of a broader strategic alliance between Recursion, Roche, and Genentech valued at up to $12 billion. The overarching partnership covers up to 40 programs across neuroscience and a GI-oncology indication. As the small molecule program advances, the industry will be watching to see if this AI-first mapping approach can consistently reduce the high attrition rates typical of neurology drug development.