Rob Knight
Director of the Microbiome Innovation Center and Professor at the University of California, San Diego (United States)
This article makes many good points. I would say that most or all of them are already well appreciated by experts in the field already, but it is useful to have them summarized in one place especially because so many people are joining the field at the moment.
The lack of replication of microbiome associations with disease is an important topic but is more nuanced than described in the article. To be useful as a clinical test, a microbe or microbiome pattern doesn’t have to cause the disease, but just act as an accurate marker for the disease. Additionally, we think it really is true that some of the markers are different in different populations, but real for each population individually. This is just like how the same drug can be more effective for one group of people than another — it doesn’t mean you shouldn’t use the drug, but it does mean you should find out who it works for rather than giving it to everyone.