Direct™ Select Technology
We are focusing our research efforts on developing and enhancing a novel drug discovery technology which we call DirectSelect™. We have made significant progress on the development of DirectSelect™, including solving many of the technical challenges associated with creating and screening ultra-large advanced combinatorial chemistry libraries. The enhanced capabilities of these libraries stem from their sheer size, up to 10,000 times the size of compound libraries typically used in the pharmaceutical industry. We believe that our DirectSelect™ technology, through the creation of these vast libraries of drug-like molecules, will allow us to more rapidly and directly identify orally available compounds with high affinity and specificity than has routinely been possible using traditional drug discovery methods. Traditional drug discovery methods often utilized by large pharmaceutical companies typically start with up to one million molecules to find a promising molecule. Subsequent optimization of that molecule to develop a lead compound may take four to five years. Our vision for DirectSelect™ is to create libraries with such high diversity -- tens’ of billions of compounds -- that the normal time frame for drug discovery can be dramatically shortened by rapidly identifying lead molecules without a significant investment in medicinal chemistry optimization. In addition, we believe libraries of this scale will yield multiple series of lead molecules, thereby providing the opportunity to select the least toxic molecules for advancement into the clinic.
During 2005, we expanded our DirectSelect™ library collection to greater than five billion compounds. The design of the libraries has been focused on creating novel, drug-like structures in accordance with various rules and guidelines used in the industry. We have blended the rules for oral availability developed by Lipinski (Advanced Drug Delivery Reviews 23, 3-25, 1997) and Veber (J. Med. Chemistry 45, 2615-2623, 2002) and others into the design of its libraries for the identification and generation of pharmaceutically relevant compounds. The design of the libraries also emphasizes occupying unpatented chemical space so that the compounds that result from the screening of the library have apparent freedom-to-operate. This issue is increasingly challenging in the industry, but one which we believe can be addressed in advance with the initial design of our libraries. In addition, we have developed a range of new chemical reactions that will allow for continued synthesis of diversified libraries.
One of the challenges that we faced during 2005 was demonstrating our ability to routinely screen libraries of millions or billions of compounds and identify novel compounds. During the fourth quarter of 2005, as a proof-of-concept, four targets were screened to identify hit molecules. These targets included three kinase and one protease target, including the cancer targets Aurora A kinase and Gleevec-resistant mutant Abl T315I kinase, the inflammation target p38 map kinase, and the Alzheimer’s target beta secretase. In each case, novel advanced "hit" molecules were identified. A “hit” is a molecule that was re-synthesized without accompanying tags used in the library version of the molecule, confirmed structurally, and shown to inhibit the biochemical activity of the target. An “advanced hit” is an identified group or family of related structures that share a relationship between structure and activity. The process of performing the screen, analyzing the output data, re-synthesizing the compounds and testing them in enzyme inhibition and cell assays took approximately one month per target. Several of the advanced hits are currently being optimized to identify “lead” molecules through the creation and screening of focused libraries using DirectSelect™. “Lead” molecules are compounds that have improved potency at inhibiting the target of interest with defined selectivity for the target. Given our internal validation of DirectSelect™'s utility in identifying hit molecules, we have initiated activities to screen for new compounds against therapeutic targets of interest. The continued development and enhancement of DirectSelect™ is a key component of our strategic operating plan, and we believe that this technology may be an important tool for the future of drug discovery and development. W e have filed United States and foreign patent applications covering the DirectSelect™ process and libraries generated under this process.
In addition to our internal efforts to develop and enhance DirectSelect™, our strategic operating plan includes entering into drug discovery and development partnerships utilizing DirectSelect™. To this end, in March 2006, we entered into a technology transfer and option agreement with Gilead Sciences, Inc. Under the agreement, we will utilize DirectSelect™ to identify lead drug candidates for an antiviral target supplied by Gilead. The companies will collaborate to identify lead molecules that have defined activities against the antiviral target. Each company will fund its own resources during the course of the research collaboration. We will own any lead molecules identified during the course of the research collaboration. Gilead has an exclusive option to enter into a license agreement for any lead molecules so identified.
We continue to initiate and advance research collaboration discussions with major pharmaceutical companies with respect to DirectSelect™. We also intend to use this technology to expand our own proprietary development pipeline.