Aureka raises $100 million: laboratory results will feed back into model training for drug discovery

Aureka raises $100 million: laboratory results will feed back into model training for drug discovery

On 10 August, biotechnology company Aureka closed a $100 million Series B funding round. The company says the funds will support drug discovery models and experiments whose results will be used in the next round of training.

Drug discovery begins with a candidate molecule, which is a potential form of a future drug. A model can propose such a candidate, while the laboratory tests three separate properties: whether it binds to the biological target, meaning the protein or other part of the organism that the drug is intended to affect; whether it performs the required function; and whether it is suitable for further development. Each measurement provides the model with a separate signal.

According to Aureka, the model proposes a molecular candidate or a testable hypothesis. The laboratory collects data on the candidate's properties. These data are then fed back into both the general model and the version used for the specific development program. The experimental result informs the next selection of candidates.

In July, Aureka released OpenDDE, a model that predicts how biomolecules assemble into a complex and interact. That release addressed a different question, such as whether an antibody can bind to a target protein. According to the company, the new funding round will support the next part of the same process: laboratory measurements of a candidate's function will be fed back into the design process.