Automated refinement and inference of analytical models for metabolic networks

Abstract

The reverse engineering of metabolic networks from experimental data is traditionally a labor-intensive task requiring a priori systems knowledge. Using a proven model as a test system, we demonstrate an automated method to simplify this process by modifying an existing or related model – suggesting nonlinear terms and structural modifications – or even constructing a new model that agrees with the system’s time-series observations. In certain cases, this method can identify the full dynamical model from scratch without prior knowledge or structural assumptions. The algorithm selects between multiple candidate models by designing experiments to make their predictions disagree. We performed computational experiments to analyze a nonlinear seven-dimensional model of yeast glycolytic oscillations. This approach corrected mistakes reliably in both approximated and overspecified models. The method performed well to high levels of noise for most states, could identify the correct model de novo, and make better predictions than ordinary parametric regression and neural network models. We identified an invariant quantity in the model, which accurately derived kinetics and the numerical sensitivity coefficients of the system. Finally, we compared the system to dynamic flux estimation and discussed the scaling and application of this methodology to automated experiment design and control in biological systems in real-time.

Half dozen of one, six billion of the other: What can small- and large-scale molecular systems biology learn from one another?

“…(It is a delightful irony that the group used genetic algorithms to make these discoveries.) Applications to biology may yield new biological laws we may never have envisioned otherwise (Schmidt et al 2011). Or perhaps we may draw inspiration from advances in computer vision, in which very large data sets coupled with large neural networks have led to stunning advances in the ability of computers to parse natural images (Deng et al 2009;Russakovsky et al 2014), with these programs now able to identify objects in images with startling accuracy.…”

Systems biology of complex symptom profiles: Capturing interactivity across behavior, brain and immune regulation

“…These examples represent an important step forward in that we are now examining illness in the context of altered regulatory circuitry. More ambitious and detailed models are now being constructed that extend the identification of the type of input-output associations to more complex model forms such as formal Hill kinetics (Schmidt et al, 2011). Similarly, added regulatory structure is also being applied to this analysis by re-organizing these networks into assemblies of formal feedback and feed-forward control elements or motifs (Alon, 2007).…”

“…At one end of the spectrum, large-scale methods are being developed to recover organism-wide metabolic regulatory reaction kinetics (Schmidt et al, 2011) from experimental data. At the opposing end of the spectrum, Craddock et al, (2012) used computational molecular dynamics combined with pharmacokinetic modeling to describe β-amyloid-induced alterations in zinc concentration and its potential effects on neuronal microtubule stability and the molecular dynamics of cognition in Alzheimer’s disease.…”

Real-Time Digitization of Metabolomics Patterns from a Living System Using Mass Spectrometry

“…Evidence suggests metabolite abundance exhibits phase like behavior when studied over time 1 . This is best described in yeast, where glycolytic and other general metabolite oscillations are well documented 2 .…”