"Cellular signaling networks have evolved to enable swift and accurate responses, even in the face of genetic or environmental perturbation. While we can readily assess dynamics in phosphorylation sites, our ability to model and predict the associated networks of kinases are hampered by the fact that we lack information on catalytic specificity for around 60% of the 538 human protein kinases (kinome). This translates into an even bigger gap in kinase-substrate relationships, where a phosphorylating kinase is only known for 20% of all known phosphorylation sites. The importance of closing these gaps is underlined by the fact that kinases are the target of about 75% of current world-wide drug development programs, and it is increasingly evident that they must be targeted in combinations, as elucidated by network models.While genomic studies are revealing large numbers of mutations in kinases in most cancers, algorithms that can assess which of these are important for tumor growth and disease progression are missing. Thus, there is a critical need for algorithms that can predict how such lesions affect the catalytic specificity of kinases. These challenges must be resolved before we can predict how combinations of genetic alterations affect networks and thereby drive complex phenotypes and diseases.The main objective of this grant is to explore the specificity space of kinases through a combination of experimental and computational approaches. We shall investigate how specificity in cellular signaling systems may be altered during both natural evolution and cancer development. We will develop a new generation of network biology algorithms to enable interpretation of mutations in the kinase domain. In combination with semi-automated specificity and mass-spectrometry interaction screening of hundreds of kinases, we shall deploy these algorithms to specifically identify drift in natural selection of kinase specificity as well as in fast evolving cancer genomes."