Our ICBP Center focuses on creating predictive models for cancer defined in terms of cellular modules (such as pathways). We will focus specifically on a class of modules that are arguably the most critical in establishing and maintaining cancer cell survival — the kinases. It is likely that all cancers have at their root the activation of one or more kinases, and therefore targeting activated kinases represents an attractive therapeutic strategy. In addition, we will focus on kinases because suitable reagents (e.g., antibodies, constructs) are available to functionally validate the models generated by this project. The lessons learned from modeling kinase activity, however, are expected to extend to other aspects of cancer systems biology.
The aim of our ICBP Center is to create computational models able to predict kinase activity in a biological sample (cell line or tumor). Focusing initially on each of the approximately 100 tyrosine kinases, the predictive models will consist of ‘cellular signatures’ (distinctive sets of co-expressed genes) that can accurately predict:
- whether a given kinase pathway is activated in the sample;
- whether a given kinase pathway is essential for viability; and
- whether a given chemical or drug modulates the kinase pathway.