Public Release: 

New computational strategy reveals genes that may help microbes adapt to the gut

Novel approach could help pave way to medical treatments that target specific genes in gut microbes

PLOS

IMAGE

IMAGE: An illustration of an association between gut prevalence and gene presence across phylogenetically related microbes. In this example, the microbes carrying a particular gene have higher gut prevalences, while their... view more 

Credit: Patrick Bradley

Microbes living in a person's gut can impact their health, but it is unclear why certain microbes colonize the gut and others do not. Specific genes may help microbes live in the gut but are difficult to identify because closely related species share many genes and tend to be present in the gut together. Thus, many genes that seem associated with living in the gut do not actually influence adaptation to this environment.

To identify microbial genes found specifically in the gut while accounting for common traits and ancestry, Bradley and colleagues employed a computational technique known as phylogenetic linear modeling. Often used in ecology studies, this approach had never before been applied to metagenomic data. The researchers used it to analyze genetic sequencing data from hundreds of people's gut microbiomes.

The analysis identified thousands of microbial genes that are over-represented in bacteria prevalent in the gut, and in bacteria that are preferentially found in the gut relative to other parts of the human body, all while accounting for evolutionary relationships between bacteria. It also uncovered genes associated with bacteria that are more prevalent in patients with Crohn's disease than in healthy controls.

"Experimentally identifying microbial genes involved in host colonization is not only labor- and time-intensive, but also requires genetic tools that haven't been developed for most gut commensals, making computational approaches particularly important," Bradley says. "This study shows that using phylogenetically-aware methods, we can predict these genes with much greater accuracy than standard computational methods allow."

The research team is now developing a tool to allow other scientists to upload their own data and use the same technique. Ultimately, the researchers hope their work will improve understanding of the gut microbiome so that it can be altered by targeting specific microbial genes. Such approaches might include new drugs to treat diseases like Crohn's, which has been linked to a loss of bacteria with anti-inflammatory effects, or to prevent colonization by harmful microbes like C. difficile.

###

Peer-reviewed Simulation / modelling Cells

In your coverage please use this URL to provide access to the freely available article in PLOS Computational Biology: http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1006242

Citation: Bradley PH, Nayfach S, Pollard KS (2018) Phylogeny-corrected identification of microbial gene families relevant to human gut colonization. PLoS Comput Biol 14(8): e1006242. https://doi.org/10.1371/journal.pcbi.1006242

Image Caption: An illustration of an association between gut prevalence and gene presence across phylogenetically related microbes. In this example, the microbes carrying a particular gene have higher gut prevalences, while their close phylogenetic 'neighbors' lack the gene and have lower prevalences.

Image Credit: Patrick Bradley.

Image Link: https://plos.io/2AARenA

Funding: Funding for this research was provided by NSF grants DMS-1069303 and DMS-1563159, Gordon & Betty Moore Foundation grant #3300 (moore.org), and institutional funds from the Gladstone Institutes (gladstone.org). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declared that no competing interests exist.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.