What evidence is there that using data to design drugs actually works?

We believe the only way to demonstrate that our system works is prospectively: in the lab, solving real industry-relevant problems. The following examples illustrate the advantages of using our technology platform. The work described was either conducted internally or in collaboration with our academic and commercial partners.

Complex multi-target profiles. We applied our technology to design compounds that would be dual-target inhibitors capable of both lowering cholesterol (inhibiting HMG-CoA Reductase) and triglycerides, as well as reducing vascular inflammation (inhibiting p38A MAP kinase). Our patented compounds demonstrate this profile in vivo.

Fast-follower programs. Our system starts with standard SAR as input but can also start from published patents. We have worked on two different fast-follower projects in HCV: NS3 Protease and NS5B. In both cases we delivered novel, potent compounds within the initial 15 compounds synthesized.

Challenging selectivity problem. We have worked on both GPCR and kinase selectivity problems. These included finding new selective alpha-1 adrenergic receptor antagonists, as well as designing a novel family of subtype-selective kinase inhibitors for cancer in collaboration with a biotech partner.

Emerging targets. Often new targets have very limited data—sometimes just 10–20 examples of weakly active compounds. In collaboration with Stanford University, we have successfully designed new families of inhibitors against transglutaminase-2 for the treatment of Celiac disease. We also have had success in a project where the goal is to design inhibitors of RNA-dependent protein kinase (PKR), an emerging target for metabolic disease.

Phenotypic starting points. We designed a novel NNRTI for the treatment of HIV, using only whole-cell replication data as a starting point. Within six months and 21 compounds synthesized, we obtained a family of inhibitors that showed broad-spectrum activity and improved pharmaceutical properties over the market leader (Sustiva®).

Crowded IP spaces. We were able to, in only a few months, identify 10 families of inhibitors of p38A MAPK. The most interesting compound proved to have potent oral activity in an in vivo model of inflammation.