Rapid computational identification of the targets of protein kinase inhibitors

This review focuses primarily on computational methods for predicting the selectivity of small-molecule inhibitors of protein kinases. A detailed discussion is presented of two computational studies that have attempted to make inhibitor selectivity predictions on a kinome-wide scale, and studies describing other methodologies that might potentially be applied to this area are also outlined. As the availability of reliable experimental data measuring inhibitor binding affinities (as opposed to the degree of inhibition of the kinase reaction) is important for the development and validation of most computational methods, recent advances in the large-scale experimental acquisition of selectivity data are also discussed, and a comparison of recently published computational selectivity predictions against a large-scale experimental screen is provided.