to BioTechniques free email alert service to receive content updates.
Revising metabolic reconstructions

05/12/2011
Julie Manoharan

A new computational model boosts the confidence of comparisons between the genome-scale metabolic reconstructions of different species.

Bookmark and Share

University of Virginia researchers have developed a new computational model that can identify and compare functional variation in the genome-scale metabolic reconstructions (GENREs) of multiple species. This model could provide new insights into the metabolic biology of genetically related species.

The computation model—called metabolic network reconciliation—removes artificial differences that are introduced into a GENRE by the reconstruction process, leaving only the biologically relevant genomic and metabolic annotations in the new reconstructions that can easily be compared with confidence. In a paper published in PLoS Computational Biology, the researchers used this reconciliation process to compare two related bacteria, Pseudomonas aeruginosa and Pseudomonas putida.

Metabolic network reconciliation removes artificial differences that are introduced into a GENRE by the reconstruction process, leaving only the biologically relevant genomic and metabolic annotations in the new reconstructions. Source: PLoS Computational Biology

“We realized that there were a lot of discrepancies between the two GENREs, and we couldn’t confidently say whether the differences were actually biologically relevant or if they were just artifacts of the reconstruction process,” says Jason Papin, assistant professor of biomedical engineering at the University of Virginia and an author on the paper.

Because GENREs integrate the genome sequence and metabolic annotations of an organism from a multitude of sources, including databases and primary literature, researchers have had varying levels of confidence about the differences they find. To increase this confidence, Papin’s reconciliation process pools all the known reactions from both species and determines whether or not the reactions should be included, based on all available information from existing annotation databases.

P. aeruginosa and P. putida have piqued the interest of the biotechnology industry because they can degrade environmental contaminants such as methylbenzene, a by-product of the petroleum industry. But while P. aeruginosa is a human pathogen, P. putida is not. The reconciliation of these species could lead to the identification of several metabolic features that may contribute to pathogenicity, including an unexpected difference: P. aeruginosa can thrive on sulfur provided by mucus-forming proteins better than P. putida.

“In the end, we’re left with two metabolic GENREs for this pathogen and nonpathogen for which every observable difference has been examined and verified with genomic, physiological, or literature-derived evidence,” Papin says. “So any comparisons we make, we can have confidence that any differences reflect true biological differences in metabolism and not reconstruction-based noise.”

The paper, “Reconciliation of genome-scale metabolic reconstructions for comparative systems analysis,” was published 31 March 2011 at PLoS Computational Biology.

Keywords:  genomics