feed

[GoogleCodeExporter]

From group#4 at Phenotype RCN meeting.

- plant and animal feed have very user-unfriendly definitions
- animal feed definition is too restrictive - mentions livestock
- we think "plant feed" is a funny label
- a grouping class for these (ie generic feed) may be useful.

My own opinion:

We should rename as "foodstuff" under a general "stuff" category. We should 
have an axiom definining this using an external role ontology. E.g. foodstuff = 
stuff and has_role food (this would be hidden from the average user). 
Subclasses could be defined appropriately using a taxonomy.

Original issue reported on code.google.com by cmung...@gmail.com on 22 Feb 2014 at 1:19

[pbuttigieg]

@Public-Health-Bioinformatics @marieALaporte @cmungall

@marieALaporte and I discussed where the various "feeds" should live and are settling on FOODON. The main reason is that FOODON will be most suited to trace nutritional contents. AGRO would contain the processes for cultivating or composting the feed material. ENVO can come in when there's some out-of-domain classes like some sort of intermediate or waste that are not in the scope FOODON/AGRO.

@marieALaporte noted that the Trait Ontology does have things like vitamin concentrations. Unsure on how to related this to FOODON. Perhaps the trait should be more neutral relative to nutrition? That is, a trait would be something like:

• the content of retinol , carotenes alpha-carotene, beta-carotene, gamma-carotene, or beta-cryptoxanthin

FOODON would then concern itself with, say, the bioavailability of these compounds in a given product?

Thoughts?

[Public-Health-Bioinformatics]

Bioavailability is a good term. I think we need a distinction between what an organism can provide itself (camel lives off of its own milk say) vs what an organism can consume from another? So FoodOn is concerned with the bioavailability of food for primary, and secondary (and tertiary) consumers. Ideally FoodOn would avoid describing the metabolic modelling of any particular food or consumer organism. FoodOn would try to describe the potential bioavailability of nutrients from a food organism. This bioavailability is direct (i.e. from ingesting raw food) or eventually taking into account some transformative process (e.g. cooking, or external digestion via enzyme).

I do see in Plant Trait Ontology for example:
beta carotene
is_a cyclic carotene
has role some antioxidant
has role some plant metabolite
has role some mouse metabolite
has role some human metabolite
has role some provitamin A
has role some biological pigment

That sounds fine to describe in TO. Whether beta carotene can be extracted from a class of organism, or reused via digestion by a class of consumer (or a particular oddball one) may be questions that FoodOn might be able to offer some (relation) generalizations about. @pbuttigieg I presume this is how one starts to get into food web relationships?

[cmungall]

On 13 Mar 2017, at 6:48, Pier Luigi Buttigieg wrote: ***@***.***ALaporte](<https://github.com/marieALaporte>) noted that the Trait Ontology does have things like vitamin concentrations. Unsure on how to related this to FOODON. Perhaps the trait should be more neutral relative to nutrition? That is, a trait would be something like: * the _content_ of retinol , carotenes alpha-carotene, beta-carotene, gamma-carotene, or beta-cryptoxanthin FOODON would then concern itself with, say, the bioavailability of these compounds in a given product?

Do breeders breed for bioavailability or just content? I think in general users of TO may care for nutritionally relevant traits.

[cmungall]

The wacky has-roles metabolites come from CHEBI. Need to get them to fix this

…

On 13 Mar 2017, at 10:51, Damion Dooley wrote: Bioavailability is a good term. I think we need a distinction between what an organism can provide itself (camel lives off of its own milk say) vs what an organism can consume from another? So FoodOn is concerned with the bioavailability of food for primary, and secondary (and tertiary) consumers. Ideally FoodOn would avoid describing the metabolic modelling of any particular food or consumer organism. FoodOn would try to describe the potential bioavailability of nutrients from a food organism. This bioavailability is direct (i.e. from ingesting raw food) or eventually taking into account some transformative process (e.g. cooking). I do see in Plant Trait Ontology for example: cyclic carotene has role some antioxidant has role some plant metabolite has role some mouse metabolite has role some human metabolite has role some provitamin A has role some biological pigment That sounds fine to describe in TO. Whether cyclic carotene can be extracted from a class of organism, or reused via digestion by a class of consumer (or a praticular oddball one) may be questions that FoodOn might be able to offer some (relation) generalizations about. @pbuttigieg I presume this is how one starts to get into food web relationships? -- You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub: #96 (comment)

[Public-Health-Bioinformatics]

Ah, I see, the mouse/human metabolites are irrelevant within PO.
Seems like Trait Ontology domain does intersect with FoodOn in a big nutritional way.