The expression of an individual gene in a developing organ is not a soliloquy; rather, it acts in a chorus of quantitative functional relations appropriately termed a genetic circuit or network. It is the task of systems biology to quantitatively define and analyze the parts (subcircuits) of the whole, the goal being to put it together in the future. Two effective strategies have emerged: (1) infer biologic pathways from large data sets and increasingly powerful tools for managing and searching literature and pathway databases; (2) construct and test increasingly complex mechanistic models of biologic systems. Our laboratory combines these strategies to determine the way in which a mechanistic pathway diagram (derived from our prior studies, the literature, and pathway databases) can be programmatically transformed to a corresponding system of ordinary differential equations. Anyone who uses such equations to model a single gene’s multiple kinetic parameters soon realizes that large network models quickly become prohibitively complicated. Thus, we find it helpful to begin with a “softfocused” level of description for a genetic network, namely to concentrate on the system behavior of the network while neglecting dynamic molecular details whenever possible. One method of achieving this utilizes Probability Neural Network modeling to derive a gene expression signature that distinguishes among physiologic states or phenotypes. Informed by these results, one may then do kinetic modeling of the complex dynamic system.

