Relative frequency of specific ASV


I have conducted a differential abundance analysis of some greenhouse samples using LEfSE and DEICODE based on two groupings: healthy and diseased. However, these two groupings come from samples obtained from 12 different greenhouses (8 were healthy and 4 were diseased). The results from both LEfSE and DEICODE pointed out a specific ASV present in samples obtained from diseased greenhouses. The sequence of the ASV found matches perfectly with that of the pathogen of interest.

Now, a reviewer is asking if that specific ASV is only present in the healthy or diseased plots. Based on a rarefied ASV table (the same one used for LEfSE), that specific ASV (which is supposed to be the causal pathogen) is present throughout most of my samples, both healthy and diseased (see graph below).

I was wondering if it would be possible to hold the hypothesis that the specific (pathogenic) ASV is more prevalent in samples obtained from diseased greenhouses even though it has been detected in other healthy greenhouses?

Thanks a lot for your input!


Hi @Pablo_V,

I think your reviewer is focusing on a Koch’s postulates.

So, then, my questions for you are whether Koch’s postulates are universally true and if they are, what else might cause contamination. I think looking into something like C. difficile infections is a good place to start with whether Koch’s postulates are always true, at least as a human disease.

I might also look into index hopping and well to well contamination as technical issues.