Modeling TMM networks in tumors

In subproject SP3 we develop mathematical network models that can rationalize the data derived from the consortium in an iterative cycle between experiments and modeling. Specifically, three goals will be reached: (i) We will develop an analysis pipeline that supports stratification of tumor entities into subgroups according to their Telomer Maintenance Mechanism (TMM). This will be accomplished by setting up a comprehensive machine learning system that incorporates genetic, epigenetic, transcriptomic and imaging data for glioblastoma and prostate cancer. (ii) We dissect the underlying mechanisms to identify key features and deregulated steps for the TMM in these tumor subgroups. To test the predictions from our model experimentally, we establish cell lines with a combination of a specific genetic background and/or knockdown of factors that represent the molecular signature of a specific TMM tumor subtype and analyze their corresponding cellular phenotypes. (iii) We conduct in silico knockout simulations with our gene regulatory network models for the different TMM subgroups and mimic perturbances of the system that are most efficient for TMM inhibition. A special focus will be on predicting the combinatorial effects of inhibiting telomerase and ALT (Alternative Lengthening of Telomers) while promoting telomere resection, which is investigated experimentally by the consortium partners.

Regulation of genes by transcription factors can be different in different cells types, and in particular tumor cells. These differences are investigated in respect to TMM.

Keywords: Gene regulatory network, systems biology, machine learning, functional genome analysis