Frobenius Discovery

We search ultra large libraries for hit-like chemical matter using statistically robust proprietary scoring functions. Probabilistic algorithms exploit the structure of combinatorial libraries, and facilitate virtual screening of libraries consisting of trillions of diverse compounds. The output of our screen is a diverse set of compounds in energetically favourable conformations hyper-enriched in key protein-ligand interactions.

We have successfully applied covalent fragment screens to find novel hit matter for previously undrugged protein families and target classes. The vast quantities of data generated by this technique maximises the speed of downstream development.

We then design and score novel synthetically accessible small molecules on a scale that would not be feasible for a team of chemists. Probabilistic modeling allows for principled