Display Settings:

Format

Send to:

Choose Destination
See comment in PubMed Commons below
Bioinformatics. 2011 Feb 15;27(4):541-7. doi: 10.1093/bioinformatics/btq702. Epub 2010 Dec 20.

Functional integration of a metabolic network model and expression data without arbitrary thresholding.

Author information

  • 1Department of Biomedical Engineering, University of Virginia, Charlottesville, VA 22908, USA.

Abstract

MOTIVATION:

Flux balance analysis (FBA) has been used extensively to analyze genome-scale, constraint-based models of metabolism in a variety of organisms. The predictive accuracy of such models has recently been improved through the integration of high-throughput expression profiles of metabolic genes and proteins. However, extensions of FBA often require that such data be discretized a priori into sets of genes or proteins that are either 'on' or 'off'. This procedure requires selecting relatively subjective expression thresholds, often requiring several iterations and refinements to capture the expression dynamics and retain model functionality.

RESULTS:

We present a method for mapping expression data from a set of environmental, genetic or temporal conditions onto a metabolic network model without the need for arbitrary expression thresholds. Metabolic Adjustment by Differential Expression (MADE) uses the statistical significance of changes in gene or protein expression to create a functional metabolic model that most accurately recapitulates the expression dynamics. MADE was used to generate a series of models that reflect the metabolic adjustments seen in the transition from fermentative- to glycerol-based respiration in Saccharomyces cerevisiae. The calculated gene states match 98.7% of possible changes in expression, and the resulting models capture functional characteristics of the metabolic shift.

AVAILABILITY:

MADE is implemented in Matlab and requires a mixed-integer linear program solver. Source code is freely available at http://www.bme.virginia.edu/csbl/downloads/.

PMID:
21172910
[PubMed - indexed for MEDLINE]
Free full text
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for HighWire
    Loading ...
    Write to the Help Desk