New relation request: has substrate

[cmungall]

proposal outlined here
https://docs.google.com/document/d/1QMhs9J-P_q3o_rDh-IX4ZEnz0PnXrzLRVkI3vvz8NEQ/edit

[phismith]

3 Introduce new sub-relation has-substrate

"p has substrate c iff p has-input c, and..."
Characteristics: Functional (a process cannot have two substrates)

Suggestion:

"p has substrate c
= def.
#p realizes function f at t & f has material basis c at t

[dosumis]

Characteristics: Functional (a process cannot have two substrates)

Not sure about this. Where does it come from? When a catalysed reaction has two inputs, how do we decide which is the substrate?

My favoured definition of has_substrate (not completely formalised).

p has_substrate c iff p is a catalytic process and c is transformed or destroyed as a result of p

Potential advantages:
clear; easy to apply; reasonable, based on usage of term (although see below).

Potential disadvantages (from discussion here: https://docs.google.com/document/d/1QMhs9J-P_q3o_rDh-IX4ZEnz0PnXrzLRVkI3vvz8NEQ/edit#bookmark=id.m04z9f7hl36d)

• Too narrow?: Doesn't cover a common use of substrate to denote the substance transported by a transporter.
  OBJECTION: I think the narrower usage I'm suggesting is reasonable allows us to be more precise.

• Too broad?: In a Protein Kinase reaction, ATP is destroyed and the protein is transformed. Under some usages of 'substrate' only the protein would be considered a substrate.
  OBJECTION: I'm not sure we can formalise this distinction. I think it is not unreasonable to call both substrates. I'd be interested to get a biochemist's take vs a molecular biologist's on the protein kinase example.

[cmungall]

On 13 Jul 2018, at 4:20, David Osumi-Sutherland wrote: > Characteristics: Functional (a process cannot have two substrates) Not sure about this. Where does it come from?

I think it has some favorable characteristics but could go either way If we drop this, does it propagate to parts? E.g. if C is a complex and a substrate, are the protein parts substrates, sometimes/always?

My favoured definition of has_substrate (not completely formalised). p has_substrate c iff p is a catalytic process and c is transformed or destroyed as a result of p

The disjointness between MF and BF is a bit awkward here, if we want this to be used e.g. catabolic processes and kinase activity

[cmungall]

Was responding to @dosumis without seeing @phismith's comment.

Barry suggests:

"p has substrate c =def p realizes function f at t & f has material basis c at t"

From my reading of section 3.7.8 in BFO2 reference this doesn't correspond to the intent of this relation, but it may help to have function elucidations in 3.7.8 as well as 'plain' dispositions.

Note: I have an outstanding ticket seeking clarification on has-material-basis on the BFO tracker: BFO-ontology/BFO#113 (was sidetracked a bit by discussion of fields, which is a different matter)

[cmungall]

I asked for an elucidation here: BFO-ontology/BFO#219 - I think this one is for you @phismith

[LEHunter]

I am confused by Barry's suggestion, "p has substrate c =def p realizes function f at t & f has material basis c at t.” Don’t all the participants together form the material basis for the realized function? Even if one were going to pull out a single material basis for realizing the function, I would think it would be the catalyst, not the (intuitively understood) “substrate".

[LEHunter]

With regard to David O-S's narrowness concern, that "p has_substrate c iff p is a catalytic process and c is transformed or destroyed as a result of p” excludes entities transported, one could either define a separate relation (has_cargo?) or refine the definition to include "...and c is transformed, translocated, or destroyed as a result..." If the latter, one could define sub-relations has_cargo and has_reaction_substrate if that would be useful.

[LEHunter]

With regard to David O-S's broadness concern, that both the kinase and the ATP are substrates of a protein kinase reaction, I agree that this doesn't bother me much. However, if one wanted to hew more closely to biologists' common usage, one could create a defined class of either ubiquitous_substrates (perhaps a cardinality restriction) or energy_contributing_substrates and exclude them from the range of the relationship. Seems unwieldy and probably unnecessary, but could be done.

[cmungall]

Don’t all the participants together form the material basis for the realized function?

I'm not sure - certainly from an intuitive interpretation of the term "material basis", all participants are both material and required. But the BFO docs explain material bases purely in terms of dispositions (AFAICT).

Even if one were going to pull out a single material basis for realizing the function, I would think it would be the catalyst, not the (intuitively understood) “substrate".

+1

[cmungall]

See PR for two sub-relations of has-input and has-output #291

This lessens the need for an explicit has-substrate relation

[wdduncan]

This issue is being closed because it has not had any activity in the last 2 years. If you think it is still worth pursuing, please reopen the issue and comment on its status.