KOPTIC: A novel approach for in silico prediction of enzyme kinetics and regulation

Kinetic models of metabolism (kMMs) provide not only a more accurate method for designing novel biological systems but also characterization of system regulations; however, the multi- omics data required is prohibitive to their development and widespread use. Here, we introduce a new approach named Kinetic OPTimization using Integer Conditions (KOPTIC), which can circumvent the omics data requirement and semi-automate kMM construction using in silico reaction flux data and metabolite concentration estimates derived from a metabolic network model to return plausible reaction mechanisms, regulations, and kinetic parameters (defined as reactomics) using an optimization-based approach. As a benchmark for the performance of KOPTIC, a previously published, four-tissue (leaf, root, seed, and stem) metabolic model of Arabidopsis thaliana was used, consisting of major primary carbon metabolism pathways, named p-ath780 (1015 reactions, 901 metabolites, and 780 genes). Data required for KOPTIC was derived from an Arabidopsis lifecycle of 61 days. Nine separate regulator restriction sets (allowing multiple solutions) defining KOPTIC runs hypothesized 3577 total regulatory interactions involving metabolic, allosteric, and transcriptional regulatory mechanisms (with nearly 40 verified by existing literature) with a median fit error of 13.44%. Flux rates of most KOPTIC fits were found to be significantly correlated with (93.6% with p < 0.05) and approximately 1:1 (r = 0.775, p << 0.001) to the input time-series data. Thus, KOPTIC can hypothesize maps the regulatory landscape for a specific reaction, out of which the most relevant regulatory interaction(s) can be defined by the desired growth/stress conditions or the desired genetic interventions for use in the creation of kMMs.