Lactobacillus in human intestine. Credit: Getty Images Plus

An algorithm can single out minimal microbiomes, the smallest groups of microbes, from a large healthy microbiome. These tiny clusters have species executing specific tasks to keep the gut healthy1.

The algorithm — minMicrobiome — could help design personalised treatment strategies for gut microbiome-related diseases, for example by reintroducing lost gut flora in some cases of severe inflammatory bowel disease and Clostridium difficile infections. It could also optimize industrial processes such as chemical production and wastewater treatment, say scientists at the Robert Bosch Centre for Data Science and Artificial Intelligence at Indian Institute of Technology Madras in Chennai.

To demonstrate the algorithm, the team looked at butyrate production of the large microbial community. Butyrate, a product of dietary fibre digestion, is a short-chain fatty acid with anti-carcinogenic and anti-inflammatory properties.

Starting with a complete set, the algorithm removed organisms one by one to find the smallest subset that can break down dietary fibre to produce butyrate.

The goal was to identify species that are essential and those that can be removed without losing desired function, says co-author Karthik Raman.

The team tested the algorithm on a synthetic microbial community and three model gut communities. When applied to the synthetic nine-member community, which included four butyrate and five non-butyrate producers, the algorithm picked Escherichia coli, Bifidobacterium adolescentis and Eubacterium rectale as the necessary minimal microbiome. The algorithm also detected the species with the highest contribution to butyrate production.

“It can home in on different minimal microbiomes if present and report their growth rate and metabolite production,” says Raman.