Description of the environment

[gdadade]

MIxS and MGnify both provide controlled hierarchical vocabulary for different kinds of environments. For our purposes it is essential to distinguish between biotic vs. abiotic environment as the top level distinction. This thread shall collect opinions and discussion results about this topic.

It seems that the highest level MIxS is providing won't work for us, example being that "human-gut" is part of this top level distinction, but we don't know what to do about animal gut samples.

As of right now the distinction between "Environmental" and "Host-Associated" the way MGnify does seems ok. We must consider, that "host" is usually used in a different context, so something like "biotic" vs. "abiotic" might be more appropriate.

[only1chunts]

sorry I missed the call! can I just check what you are meaning by the MIxS "controlled hierarchical vocabulary"? as we technically don't have one for MIxS. MIxS has a mandatory field for "Board-scale environment" in which we mandate the use of the Environment Ontology (browse it here ENVO ontology) which is externally controlled (with input from GSC members). Unless you are meaning something other than the content of the "broad-scale environment" field?
I suspect the branch of the ontology of most relevance to you would be the independent continuant part where it lists biological, material, immaterial and anatomical as the top levels.
The folks working on ENVO are happy to discuss changes and help make it as useful as possible, so if there are real-world scenarios that need modelling I'm sure they would be open to working with you.

[gdadade]

MIxS uses different environmental packages. In those packages you have a set of terms. The combination of the name of the package (e.g. "air" or "human-gut") and the set of terms within those packages to me is some kind of a hierarchy. We found, that the environmental package names won't work for our purposes most likely.

I'll add the ENVO ontology to our overview doc.

[only1chunts]

package names are in no way a controlled vocabulary and do not represent anything like a list of environments so it's not surprising they won't work for anything other than a name for a packet of terms :-D

[gdadade]

Ah, thanks! So we don't need to worry about it :)