Anagenex and Nimbus join forces to develop small molecule treatments using AI

According to the deal terms, Anagenex will generate billions of biochemistry data points for each disease target nominated by Nimbus. Anagenex will then use these data to train AI models that can design 100 million new molecules that bind to the targets, and select the best candidates for developing into drug candidates. The therapeutic areas of interest were undisclosed.

In return, Nimbus will pay Anagenex an undisclosed amount upfront in addition to option and developmental milestone payments for each program in the collaboration.

“We are thrilled to be partnering with Nimbus, one of the original and most successful computationally aided drug discovery and development companies,” said Nicolas Tilmans, CEO of Anagenex, in a public release. “We’re very excited to join forces with them in attacking new targets beyond oncology a few hundred million compounds at a time.”

The vast majority of possible protein targets lack drug candidates because they don’t fit conventional understanding about how proteins can be targeted by small molecules, says Anagenex’s website. It is therefore difficult to find a “chemical starting point” for designing new drugs for these targets, and it can take years of trial and error to turn the starting point into a drug candidate, adds the website.

To solve this bottleneck, Anagenex uses a mixture of machine learning and tools such as DNA Encoded Libraries (DELs) and Affinity Selected Mass Spectrometry (ASMS) to design drug candidates more quickly and efficiently than traditional methods. To finance the development of its technology and build a treatment pipeline, the firm raised a Series A round worth $30 million in 2022.