Kinase Inhibitor Chemistry Charting the Chemical Space 2011

COMPUTATIONAL APPROACHES

Knowledge Based Computational Approaches and Efficient X-Ray Structure Mining Tools for Kinase Drug Design
Kiyoyuki Omoto, Ph.D., Principal Scientist, Medicinal Chemistry, Pfizer
Immense knowledge and data has been accumulated around kinase inhibitors through the 20-year investment by pharmaceutical/biotech industries in addition to a number of academic and non-profit organizations. Numerous selectivity data and hundreds of kinase X-Ray structures are available through external and internal proprietary databases. Although small amounts of data can be extracted from the databases with simple queries, it is manually unfeasible to harness the wealth of data from these databases. But through the application of various computational techniques it is possible to package different parts of the data for a synergetic use which would allow a comprehensive understanding of kinase inhibitors and their molecular interactions. The talk will describe (1) Gini coefficient as a way of quantifying kinase selectivity data, (2) a program for compilation of hinge-binding fragments of kinase inhibitors, (3) a hybrid design approach utilizing the hinge-binding fragments, and (4) automated way for annotating DFG-in and DFG-out conformations of the kinase. Effective practical usage of these programs for better compound design will be discussed.