I'm glad that cMD exists. However, I feel some datasets (especially oral and skin ones) could have better specification of the actual subsite assessed. I understand that oftentimes this info is not available, but I checked a few available studies and found some in which including a body_subsite column in the metadata is indeed possible (like the one already present in FerrettiP_2018, GhensiP_2019, and many others).
Would you be interested in a PR in the curatedMetagenomicDataCuration repo with this? I'd be happy to help.