Machine Learning for Uncertainty Reduction in Biochemical Kinetic Models

The primary goal of kinetic models is to capture the systemic properties of the metabolic networks, and we need large-scale kinetic models for reliable in silico analyses of the complex dynamic behavior of metabolism. However, parameter uncertainty hinders the development of kinetic models and uncertainty levels increase with the model size. Current methods for building kinetic models within constraint-based modeling frameworks address uncertainty indirectly by integrating data from different biological origins. We recently proposed iSCHRUNK, a computational approach which combines Monte Carlo sampling methods and machine learning techniques to characterize the uncertainties and to reveal complex relationships between the kinetic parameters and the responses of the metabolic networks. Monte Carlo sampling methods allow us to exploit synergies between different data sources and generate a population of kinetic models that are consistent with the available data and physicochemical laws. The machine learning allows us to data-mine the a priori generated kinetic parameters together with the integrated datasets and derive values of kinetic parameters consistent with the observed physiology. In this work, we modified iSCHRUNK to address a design question: can we identify the kinetic parameters and their values that give rise to a desired metabolic behavior? Such information is important for a wide variety of studies ranging from biotechnology to medicine. As an illustration, we used the proposed methodology to find parameters that ensure a rate improvement of the xylose uptake (XTR) in a glucose-xylose co-utilizing S. cerevisiae strain. Our results indicate that only three kinetic parameters need to be accurately characterized to describe the studied physiology, and ultimately to design and control the desired responses of the metabolism. This framework paves the way for a new generation of methods that will systematically integrate the wealth of omics data and extract the information necessary for metabolic engineering and synthetic biology decisions.