(398f) The Impact of Resource Constraints on the Reverse Engineering of Biological Pathways

Authors: 
Quarton, T., University of Texas at Dallas
Sontag, E., Northeastern University
Bleris, L., The University of Texas at Dallas
Every element within a gene network requires chemical species supplied by the cellular environment to enable biological processes. In many cases, these chemical species, or resources, act in an indiscriminate manner and are consequently shared within the network. This global coupling of the network elements through these resources can have measurable effects on the network as it operates. Here we employ an engineered gene circuit, which was introduced in a recent paper by the authors (Kang et al. 2016), and report additional data, analysis, and interpretation. Using this synthetic network as a benchmark, we find that network inference results can to be influenced by changes in the resources allocated to each node. Specifically, we show that applying Modular Response Analysis to this network leads to the inference of a nontrivial â??ghostâ? regulation edge which was not explicitly engineered into the network and which is, in fact, not immediately apparent from experimental measurements. Various mathematical models including a coupled system of ODEs, a multiclass queueing network, and a graph theory consideration are explored in closed-form. Taking resource availability into account during reverse engineering allows closer approximation of the complex cellular environment and potentially uncovers network operation constraints to be considered for improved synthetic circuit design.