Open Competition 1 - Information Technology

Chemical Intelligence Platform for Rapid Discovery of Drugs Leads

Sponsor: Eidogen-Sertanty (formerly Libraria, Inc.)

• Project Performance Period: 11/1/2002 - 10/31/2004
• Total project (est.): $2,674,738.00
• Requested ATP funds: $1,971,990.00

Libraria will streamline and improve the design of new chemicals by an innovative methodology that will search for novel chemical structures and select promising molecular candidates for new protocols (sequences of chemical reactions that transform reactants into intermediate and final products). A flexible chemical protocol compiler will rapidly compute novel protocols, and a suite of software tools will generate new molecules in software and specify the protocols for synthesizing them. A graphical user interface then will permit direct supervision of the compiler, simulating the knowledge of a chemist through a database of chemical transformations. A chemical intelligence platform then will integrate these elements for optimal effect in predicting the effectiveness of candidate molecular structures. Given this head start on the search, a chemist will then synthesize a reduced set of prioritized candidates and assays will be performed on them. The two-year project will attempt to refine the platform in three stages by applying it to known targets, such as the kinase gene family, that will serve as a test bed for evaluation, statistical analysis, quality control, and feature definition. By automating the selection of new chemical molecular structures through a combination of computer science, biology, and chemistry, Libraria will attempt to generate candidate molecules and predictions faster and better than current state-of-the-art techniques. The prototype software modules will mimic the intuition chemists use to simplify the screening process in searching for chemical structures with potential as quality drug leads. The high-speed protocol compiler will rapidly screen millions of possibilities, thereby reducing the time, effort, and expense involved in the research and development of new drug therapies. Because of the lack of precedent with this software, there is a significant risk that the algorithms developed will be either too narrow or too broad to generate sufficiently diverse candidate molecules that also can be synthesized. The pay-off is in enabling life-sciences companies to discover better "best-in-class" drugs while lowering R&D costs by approximately one-half, from $30 million to $15 million per candidate drug. Pharmaceutical companies could realize a 9 percent savings in total R&D costs, approximately $2.1 billion per year. The technology could stimulate greater competition among pharmaceutical companies, potentially lowering consumer drug prices. In addition, these improved information tools also have the potential to benefit the development of advanced materials such as ceramics, polymers, optical materials, coatings, and catalysts. Due to the project's high technical risk and the long lag time to produce investment returns inherent in the software element of the project, ATP funding is needed to allow Libraria to accelerate its development efforts to meet the anticipated advances in small-molecule drug discovery in the industry in the next five to eight years.

For project information:

Debra Bannister, (530) 676-8001
debban@pacbell.net

ATP Project Manager

Jack Boudreaux, (301) 975-3560
jack.boudreaux@nist.gov