Investigating Unknown Regions of E. coli Metabolism

Student Author(s)

John Peterson
Megan Oostindie

Faculty Mentor(s)

Dr. Aaron Best, Biology

Document Type


Event Date



Metabolic models are useful for a number of applications in the world of biology. If one understands the biochemistry involved in every part of an organism, it could be possible to predict how the organism would behave at the molecular level in particular environments. Once one model is understood sufficiently, the concepts could be applied to other organisms and groups of organisms. This has possible applications in medicine and environmental science concerning microbiomes. The base model could also be used as a foundation for further experimentation and metabolic modeling projects on more complex or less understood organisms. Our work focuses on the state-of-the-art metabolic model for E. coli, since this is a well studied model organism. However, the model currently only accounts for conservation of mass and the presence of pathways connecting metabolites. In order to be more useful, the model would have to incorporate a number of biological patterns. One such pattern is transcriptomics: the analysis of gene expression over the entire genome when the organism is in different environments. Transcriptomics is of particular interest as it involves, to some degree, all parts of the metabolic network. However, the usefulness of these data depend upon the diversity of the experiment pool. The current transcriptomic data set for E. coli contains about 1200 experiments, representing fewer than 100 unique conditions. We have conducted initial algorithmic development and analysis to identify parts of the metabolic network not yet perturbed in transcriptomic data. The goal of this project is to design a set of unique conditions and gather expression data. These data will be used to improve upon the initial algorithms and to eventually supplement the transcriptome data set with enough information to improve the current computer model of E. coli K12.


This research was supported by the National Science Foundation under grant No. MCB-1330734.

This document is currently not available here.