Computational systems biology of colorectal carcinoma relevant signalling pathways and their cross talk using nested effects models

Coordinator:		Prof. Dr. Rainer Spang
Institution:		Institut für Funktionale Genomik, Universität Regensburg
Homepage:		www-compdiag.uni-regensburg.de

Tumours arise from dysfunctional cellular communication. Normal cells grow when receiving growth signals. Tumour cells grow even without these signals and proliferate although they should not. They receive signals inducing cell death but do not respond to them properly. They can also escape immune responses by sending out signals that modulate the immune system. Tumour cells can do this by developing survival mechanisms which interfere with cell signalling pathways that control cell growth, cell proliferation, and cell death.
One of the most important channels of intra-cellular communication in colorectal carcinomas is the Wnt-Signalling Pathway. It is always disrupted in this type of cancer. Since the pathway controls the activity of transcription factors, it leaves traces in expression profiles forming characteristic patterns. These patterns can be recorded using microarrays and allow us to “listen” to Wnt-communication. In our NGFN-Plus project, we use this data to analyse the structure of the pathways including its cross- talk with different channels of cellular communication. Our approach is to experimentally perturb the pathway at well-defined nodes and record the downstream alterations of cellular communication caused by this stimulus. We use an in-house developed computational method called nested effects models to combine the data into a single pathway model. This is a joint project with the group of Prof. Michael Boutros in Heidelberg. While the Boutros Lab produces the data, we- as a bioinformatics group- model and integrate it.


          
    
        
  


       
Figure Legend: Downstream effects of simulated intervention experiments together with the underlying signalling pathway. The columns of the heat map to the right correspond to interventions in 10 signalling molecules named S1 to S10. The rows correspond to the expression of genes affected by the intervention. Yellow indicates a strong downstream effect, while blue indicates no effect. The network on the right encodes the flow of signal underlying the data pattern.