Noise processing by networks

Cell behaviour is determined by complex molecular regulatory networks. Signalling networks are particularly important since they are responsible for the robust transmission of noisy environmental information to the cell's nucleus. However, although important signalling pathways have been well studied, and the manner of noise propagation through regulatory networks has been discussed, the relationship between network architecture and the cell's ability to process environmental noise is not well understood. To approach this problem in this thesis we derive a mathematical formula relating a network's structure to its noise processing ability. We noticed that noise processing is highly affected by the networks complexity, and in particular by the number and length of the weighted paths from the noisy input(s) to the output. In order to explore the utility of this mathematical expression, we apply it to the regulatory network for pluripotency in mouse embryonic stem (ES) cells and assess the effects of network topology on the propagation of noise through this system. We conclude by using the underlying theory to explain the interaction patterns in the ES cell's transcriptional circuit.