Identifying Latent Phenotypes in the Design of Gene Circuitry through a Novel Modeling Approac

A fundamental challenge of modern biology is to understand how intricate interactions between biochemical components bring about rich phenotypes. The challenge is difficult mostly due to the inherent non-linearity of biochemical mechanisms; therefore, a deep understanding cannot be achieved by intuitive analysis and requires rigorous quantitative methods such as those provided by mathematical analysis. The conventional strategy used to analyze mathematical models typically begins with a nominal parameter set established on the basis of experimentally measured or estimated values. The result is unlikely to identify hidden phenotypes latent in any particular system design. This work develops a radically new modeling strategy that does not require specific values for any of the parameters and inverts many of the typical steps in the conventional, parameter-centric, approach. This novel, phenotype-centric, approach utilizes the system design space methodology based on a rigorous definition of qualitatively-distinct phenotypes. The strategy provides a global perspective on system behavior through four important innovations: (a) enumeration of the repertoire of qualitatively-distinct phenotypes for a system, (b) generation of parameter values for any particular phenotype, (c) simultaneous realization of parameter values for several phenotypes to aid visualization of transitions from one phenotype to another, in critical cases from functional to dysfunctional behavior, and (d) identification of ensembles of phenotypes whose expression can be phased to mimic that of a natural system or to rationally engineer a desired synthetic construct.