Doubt Ancom analysis

[rparadiso (Rubina)]

Hello to all!
I need to have some clarification regarding how to set the ancom test on my samples.
I have 60 samples divided into two groups (treated and control) and undergoing to two different diets. For each subject included in the study, samples were taken from three different body sites (rumen, feces and intestine). Now I need to identify features that are differentially abundant across two experimental groups (treated and control).
can someone help me to understand how to perform ancom analysis?
In particular how should I filter my data? my p-where is Diet?

qiime feature-table filter-samples \
  --i-table table.qza \
  --m-metadata-file sample-metadata.tsv \
  --p-where  Diet ???
  --o-filtered-table diet-table.qza


qiime composition ancom 
  --i-table comp-diet-table.qza 
  --m-metadata-file metadata.tsv 
  --m-metadata-category Subject ID ???
  --o-visualization ancom-Subject.qzv

Thanks in advance for your help.

[mortonjt (Jamie Morton)]

Maybe these docs provide an example close enough to what you need?
https://docs.qiime2.org/2020.8/tutorials/filtering/#metadata-based-filtering

If you want to run two ancom tests, perhaps you can create 2 tables with the following commands.

qiime feature-table filter-samples \
  --i-table table.qza \
  --m-metadata-file sample-metadata.tsv \
  --p-where  "Diet='diet1'"
  --o-filtered-table diet1-table.qza

qiime feature-table filter-samples \
  --i-table table.qza \
  --m-metadata-file sample-metadata.tsv \
  --p-where  "Diet='diet2'"
  --o-filtered-table diet2-table.qza