Study proposes new framework to identify keystone microbial species

Identifying keystone species can be challenging, especially in the human gut, since it is not feasible to isolate them through systematic elimination.

Machine-learning model trained on real human gut microbiome data from a curated metagenomic database was able to simulate the removal of any species in any gut microbiome sample.

This "thought experimen Brigham and Women's Hospital the relative essential the Mass General Brigham ach community. Brigham and Women's Hos DKI al the Channing Division of Network Medicine PhD Yang-Yu Liu DKI