New algorithm suggests four-level food web for gut microbes: Novel modeling approach could help improve understanding of human gut function – Science Daily

Restaurant News

A new computational model suggests that the food web of the human gut microbiome follows a hierarchical structure similar to that of larger-scale ecosystems. Tong Wang of the University of Illinois at Urbana-Champaign and colleagues present the model in PLOS Computational Biology.

In the human gut, hundreds of species of microbes exchange nutrients in a complex food web. Large-scale food webs, such as those of tropical forests, typically follow a hierarchy in which energy flows from plants, to herbivores, to carnivores. Wang and colleagues wondered if the gut microbiome could be considered to follow a similar hierarchy, from microbes that consume nutrients in food eaten by the human host, to those that eat nutrients produced by the first microbes, and so on.

To address this question, the researchers developed a computational model that uses the known species of microbes in a person’s gut to predict microbial metabolites — the substances the microbes generate as part of their biological activities, and which may serve as nutrients for other gut microbes. The metabolite predictions generated by the model are in line with experimental data, providing support for its accuracy.

The new model indeed predicts a four-level hierarchy for the food web of the gut microbiome. This suggests that species composition systematically changes along the length of the gut. Near the entrance to the lower gut, one might find bacteria from the highest hierarchical level — those that consume nutrients in food eaten by the human. Near the end of the gut, one might find bacteria from the lowest level.

“There is a great premium on being able to predict metabolic profiles from species genomes, as our algorithm does,” Wang says. “Metabolites are better than species composition for predicting important aspects of gut function, but genome sequencing is faster and cheaper than measuring metabolic profiles.”

The researchers are now working to refine their model by using a machine-learning approach to infer important competitive relationships between gut microbes. Doing so could improve the model’s accuracy, potentially reducing the need for expensive measurements of metabolic profiles in research on gut function.

Story Source:

Materials provided by PLOS. Note: Content may be edited for style and length.

Source: Thanks