We characterize different cell states, related to cancer and ageingphenotypes, by a measure of entropy of network ensembles, integrating geneexpression values and protein interaction networks. The entropy measureestimates the parameter space available to the network ensemble, that can beinterpreted as the level of plasticity of the system for high entropy values(the ability to change its internal parameters, e.g. in response toenvironmental stimuli), or as a fine tuning of the parameters (that restrictsthe range of possible parameter values) in the opposite case. This approach canbe applied at different scales, from whole cell to single biological functions,by defining appropriate subnetworks based on a priori biological knowledge,thus allowing a deeper understanding of the cell processes involved. In ouranalysis we used specific network features (degree sequence, subnetworkstructure and distance between gene profiles) to obtain informations atdifferent biological scales, providing a novel point of view for theintegration of experimental transcriptomic data and a priori biologicalknowledge, but the entropy measure can also highlight other aspects of thebiological systems studied depending on the constraints introduced in the model(e.g. community structures).
展开▼
