Inferring rates in biochemical networks using temporal fluctuations

Biochemical networks in cells are intrinsically stochastic due to the probabilistic nature of individual reaction events. Due to their complexity only some interactions within a larger network are known. The fact that many models may fit a given dataset poses a significant challenge when inferring reaction rates to construct system-level models of cellular processes. Previously, static snapshots of gene expression levels across clonal populations have been used to establish bounds on rates in several classes of gene expression models—independent of other dynamics in the system. We extend this analysis to temporal fluctuations of a molecular species in similar classes of gene expression models to determine the functional form of biochemical rates. This approach only requires observations of the species of interest and species that influence its dynamics directly. Temporal fluctuations contain more information than static snapshots, leading to more discriminatory power.