Issue 70 in ogms: new term - 'medication'

[og...@googlecode.com]

Status: New
Owner: albertgo...@gmail.com
Labels: Type-Defect Priority-Medium

New issue 70 by albertgo...@gmail.com: new term - 'medication'
http://code.google.com/p/ogms/issues/detail?id=70

Adrien Barton (at INSERM) has suggested that OGMS would be a good place for
the term 'medication' (also, 'medicine').

Do we agree it belongs here?

Here's a (very weak) first stab at a definition:

"A obi:processed material whose function is to alleviate the signs and
symptoms of a disease or negatively regulate the effects of a disorder and
designed to be realized during part of a treatment process"

Please help improve. In particular, we need to refine the genus somehow.

[Yu Lin]

I think so.
Same as definition of treatment. We need to find the scope and then
define this term.
Do we need to add the health provider party into the definition? Dose
it must to be described?

Drug will be its sub-class. How about vaccine?

Best,
Asiyah

[Bill Hogan]

I think it makes sense.  Not that we want to adopt a regulatory definition of drug, especially that of just one nation, but nevertheless there is fodder for discussion here:

http://www.fda.gov/cosmetics/guidancecomplianceregulatoryinformation/ucm074201.htm

I

[Bill Hogan]

How does drug differ from medication?

These terms are used interchangeably by both public and professionals alike.

Bill

[Richard Scheuermann]

Albert,

Can you explain to me why you insist on using signs and symptoms in your definitions? Medications can only function by targeting independent continuants. This may result in effects on dependent continuants and occurrents, but those are secondary effects. And your definition would not include curative treatments that eliminate disorders.

Richard

On Sep 21, 2011, at 11:35 AM, <og...@googlecode.com>
wrote:

------------------------------------------------------------------------------------------------

Richard H. Scheuermann, Ph.D.
Chief, Division of Biomedical Informatics
John H. Childers Professorship in Pathology
Department of Pathology
U.T. Southwestern Medical Center
5323 Harry Hines Blvd.
Dallas, TX 75390-9072

phone: 214-648-4115
FAX: 214-648-4070
email: richard.s...@utsouthwestern.edu
http://pathcuric1.swmed.edu/Research/scheuermann.html

________________________________

UT Southwestern Medical Center
The future of medicine, today.

[Mark Jensen]

By making medication a kind of processed material, we eliminate the possibility of non-processed materials being medications. Is it possible there could be a non-processed (raw) material which is also a medication (eg, bee venom maybe)? If all medications are "created or changed during material processing", then how would a drug be differentiated (if at all) from medication? Granted, extracting a natural substance from an animal or plant (to be used as a medication) is a kind of processing, it may be questionable that it has undergone a change unless it is refined or diluted.

Mark

[Yu Lin]

Considering that there is no clear definition of signs and symptoms
available yet,
I agree that using signs and symptoms in the definition make things
more difficult.

As an example, the definition for vaccine is :

A vaccine is a processed material with the function that when
administered, it prevents or ameliorates a disorder in a target
organism by inducing or modifying adaptive immune responses specific
to the antigens in the vaccine.

We didn't mention anything of signs and symptoms in the definition.

However, do you think that a vaccine is a subclass of medication?

A drug can be a pain killer which target to the symptom of pain, what
ever the cause of the pain is.

Best,
Asiyah

[Mark Jensen]

By making medication a kind of processed material, we eliminate the possibility of non-processed materials being medications. Is it possible there could be a non-processed (raw) material which is also a medication (eg, bee venom maybe)? If all medications are "created or changed during material processing", then how would a drug be differentiated (if at all) from medication? Granted, extracting a natural substance from an animal or plant (to be used as a medication) is a kind of processing, it may be questionable that it has undergone a change unless it is refined or diluted.

Mark

[Albert Goldfain]

Richard,

I don't insist on using signs and symptoms...just wanted to be overinclusive in the first pass.

In fact, I would prefer if we used just disorder here...mirroring whatever we end up doing with treatment (specifically, the "root of the disorder" in your characterization... that is in fact the target of the treatment).

All,

Not sure if there is a difference between drug and medication...but didn't want to assert that unless you all agree to it.

AG

On Wed, Sep 21, 2011 at 1:47 PM, Richard Scheuermann <Richard.S...@utsouthwestern.edu> wrote:

[Alan Ruttenberg]

OBI has a definition of drug role that I think is a good start (quick scan suggests it needs at least a grammatical fix). Modifications should be proposed, if necessary, submitted to OBI, and imported into OGMS. Looking it over I think that we might consider whether it would make more sense to have the drug role inhere in a bulk amount of material entity, and to see whether we also want to subclass it to add a tighter subclass for "good" drugs, i.e. excluding substances such as truth drugs. Perhaps this is the distinction captured between the words "drug" and "medication".

OBI also has 'role of regulator of consumables and medical devices'. A subclass of drug role could be something like "government approved drug role" and its definition could be constructed, in part, by reference to the above role.

From the role we can construct a material entity class "medication" synonym "drug", though my preference would be to instead build out from the drug role (which we need anyways) and advise that good practice is to assign the role to the substance rather than reiterate the role hierarchy in the material hierarchy. 

I concur with Mark that we don't want to assert processed material as there are unprocessed materials that probably satisfy. 

To answer Asiah's question, under the OBI definition, a vaccine particle would bear drug role. Vaccine role would be a subclass of drug role.

I don't think that disorder should be mentioned in the definition for the reason that Asiah mentions re:pain, at least until we have a consensus regarding the earlier discuss of pain/disorder.

Regarding the initial definition offered, I don't think that all drug realizables are functions - having a function would imply that it was designed or evolved and not all drugs are.

There is also an issue to perhaps discuss in whether/how drug roles inhere in molecular entities or in macroscopic material entities, what to say about non-drug constituent such as capsules, fillers, coatings and other components of the formulation.

-Alan

drug role

• Term IRI: http://purl.obolibrary.org/obo/OBI_0000040
• definition: a role borne by a molecular entity and is realized in a process of absorption by an organism alters, or effects (or is assumed to effect) a function(s) which inhere in an organism
• definition editor: Role Branch

• alternative term: drug
• example of usage: http://www.answers.com/topic/drug 1. A substance used in the diagnosis, treatment, or prevention of a disease or as a component of a medication. 2. Such a substance as recognized or defined by the U.S. Food, Drug, and Cosmetic Act.
• definition source: OBI, CDISC
• has curation status: metadata complete

Class: role of regulator of consumables and medical devices

• Term IRI: http://purl.obolibrary.org/obo/OBI_0000273
• definition: a regulator involved with making and enforcing legislation and governmental orders relevant to the development, testing, manufacture and use of food, drugs and medical devices

Annotations

• definition editor: Jennifer Fostel
• alternative term: drug regulatoy authority
• example of usage: FDA, EMEA; http://www.fda.gov/; The International Conference of Drug Regulatory Authorities (ICDRAs) provide drug regulatory authorities of WHO Member States with a forum to meet and discuss ways to strengthen collaboration.http://www.who.int/medicines/areas/quality_safety/regulation_legislation/icdra/en/index.html
• definition source: OBI, CDISC
• has curation status: metadata complete

[Albert Goldfain]

Thanks Alan,

This looks like a good starting point.  These are tricky terms because there is a regulatory perspective, pharma perspective, a biological perspective, and a clinical/medical perspective on what these things are.

Is there a principled way for us to decide what should go into OGMS and what should go into OBI?  I am very happy keeping OGMS as small as possible and MIREOTing terms from OBI...but it seems that "general medical science" should refer to of 'drug' and 'medication'. 

AG

[Yu Lin]

HI,

My two cents:

1. OBI use an alternative term 'drug' to refer 'drug role'.
I think that is not proper to do so.
It is better to differentiate 'drug role' and 'drug'.

2. OGMS may deal more with a licensed product which has been proved by
FDA or drugs used in the clinical trail.

In biomedical experiment, a lot of chemicals or any kinds of
molecular entity can be used as drug to test.
Sometimes it is not a final product. It may fail as to become a
drug which will be produced by pharmaceutical companies.
Sometimes in biomedical experiment, a licensed drug will be used.
In such a case, we do need a term 'drug' from OGMS imported to OBI.

I suggest that OBI will drop 'drug' as the alternative term for 'drug role'.
And OGMS will have 'drug' term which must be connected to FDA proved
process or license or something in a more proper English.

Best,
Asiyah

[Alan Ruttenberg]

On Thu, Sep 22, 2011 at 10:22 AM, Albert Goldfain <albertg...@gmail.com> wrote:

Thanks Alan,

This looks like a good starting point.  These are tricky terms because there is a regulatory perspective, pharma perspective, a biological perspective, and a clinical/medical perspective on what these things are.

Is there a principled way for us to decide what should go into OGMS and what should go into OBI?  I am very happy keeping OGMS as small as possible and MIREOTing terms from OBI...but it seems that "general medical science" should refer to of 'drug' and 'medication'. 

General principle is that if an ontology exists that covers the domain, have them define it. However some of these terms were defined before OGMS existed. An overriding concern, in my opinion, is stability of terms and I don't think that creation of an ontology is a sufficient reason for changing an identifier. However you could negotiate with OBI to take over maintenance of the term while keeping the OBI id, in which case the source would live in OGMS (with an OBI id) and OBI would import it back. Submit a tracker item to OBI if you would like to do that.

-Alan

[Alan Ruttenberg]

On Thu, Sep 22, 2011 at 11:00 AM, Yu Lin <lini...@gmail.com> wrote:

HI,

My two cents:

1. OBI use an alternative term 'drug' to refer 'drug role'.
   I think that is not proper to do so.
   It is better to differentiate 'drug role' and 'drug'.

Alternative terms have low standards for acceptance. However the appropriate way to have the discussion would be to submit a ticket to the OBI tracker raising this as an issue.

 

2. OGMS may deal more with a licensed product which has been proved by
FDA or drugs used in the clinical trail.

Well, the FDA isn't the only regulatory agency in the world. But perhaps you are right, in which case the regulated drug role could be defined in OGMS, inheriting from OBI's drug role.

 

   In biomedical experiment, a lot of chemicals or any kinds of
molecular entity can be used as drug to test.
  Sometimes it is not a final product. It may fail as to become a
drug which will be produced by pharmaceutical companies.
  Sometimes in biomedical experiment, a licensed drug will be used.
In such a case, we do need a term 'drug' from OGMS imported to OBI.

I suggest that OBI will drop 'drug' as the alternative term for 'drug role'.
And OGMS will have 'drug' term which must be connected to FDA proved
process or license or something in a more proper English.

I prefer more specific terms. Since "drug" is clearly used in a number of senses let's define (at least) "regulatory sanctioned drug role" def: a drug role that comes into existence by the act of a bearer of role of regulator of consumables and medical devices that sanctions the bearers use in certain medical contexts". (or something like that). You can have "drug" be an alternative term for that too ;-)

-Alan

[Alan Ruttenberg]

On Thu, Sep 22, 2011 at 10:22 AM, Albert Goldfain <albertg...@gmail.com> wrote:

Thanks Alan,

This looks like a good starting point.  These are tricky terms because there is a regulatory perspective, pharma perspective, a biological perspective, and a clinical/medical perspective on what these things are.

I'd say less tricky as you have already articulated what the ingredients are. Just define these separately.

[KerstinF]

Interesting thread. The BFO based Translational Medicine Ontology
(TMO) (first called 'Pharma Ontology') have structured the different
meanings of ‘drug’, by identifying different types of entities and
making relationships between them explicit. I am not updated on the
latest development but this how I understood it last year: ’Molecular
entity’ for single molecule kinds. ‘Active ingredient’ as a role
played by a chemical substance which has the disposition to treat a
certain disease and is part of a ‘pharmaceutical formulation’ for
biologically active chemicals. A ‘Pharmaceutical product’ is a
formulated pharmaceutical that has been approved.

My thinking is that it would be good if TMO could build on OGMS when
applicable - So maybe the "medication/drug" etc could be one
opportunity for this. So, I have copied in Joanna Luciano as we
briefly talked about this at ICBO this summer.

Regards
Kerstin Forsberg
AstraZeneca

[Yu Lin]

My thinking is that TMO should build on OGMS plus OBI.

PS: how can I join the TMO discussion mail list?

Best,
Asiyah

[Alan Ruttenberg]

On Thu, Sep 22, 2011 at 12:57 PM, KerstinF <kerstin.l...@gmail.com> wrote:

> Interesting thread. The BFO based Translational Medicine Ontology
> (TMO) (first called 'Pharma Ontology') have structured the different
> meanings of ‘drug’, by identifying different types of entities and
> making relationships between them explicit. I am not updated on the
> latest development but this how I understood it last year: ’Molecular
> entity’ for single molecule kinds.

Ugh. They've minted their own duplicate URI for molecular entity
instead of using CHEBI's directly. Also duplicates of a bunch of OBI
terms.

‘Active ingredient’ as a role
> played by a chemical substance which has the disposition to treat a
> certain disease

Seems too narrow. Codeine as active ingredient in a pain killer is not
necessarily used to treat any disease.

> and is part of a ‘pharmaceutical formulation’ for
> biologically active chemicals. A ‘Pharmaceutical product’ is a
> formulated pharmaceutical that has been approved.

Again formulated pharmaceutical seems too narrow. If you have an
herbal tea with medicinal properties and the molecular type is known,
would it not be considered an active ingredient?

> My thinking is that it would be good if TMO could build on OGMS when
> applicable - So maybe the "medication/drug" etc could be one
> opportunity for this. So, I have copied in Joanna Luciano as we
> briefly talked about this at ICBO this summer.

Might be a good start to first review and incorporate OBI.

-Alan
>
> Regards
> Kerstin Forsberg
> AstraZeneca

[Yu Lin]

Agree with Alan.
-Asiayh

[og...@googlecode.com]

Updates:
Status: Accepted

Comment #1 on issue 70 by albertgo...@gmail.com: new term - 'medication'
http://code.google.com/p/ogms/issues/detail?id=70

It was suggested that OGMS MIREOT the term 'drug role' from OBI for this
purpose (http://purl.obolibrary.org/obo/OBI_0000040)

[Sivaram Arabandi, MD]

Wondering if the term 'Medication' has a connotation of 'beneficial' whereas a 'drug' can be good or bad (or yet to be found out as in experimental drugs undergoing testing).

Re 'codeine', while the use of the word disease in the definition maybe too narrow, it nevertheless would be considered a 'medication' since it is used to treat a manifestation of a disease or a disorder. 

cheers

Sivaram

[Alan Ruttenberg]

On Tue, Jun 19, 2012 at 6:11 PM, Sivaram Arabandi, MD
<sivaram....@gmail.com> wrote:
> Wondering if the term 'Medication' has a connotation of 'beneficial' whereas
> a 'drug' can be good or bad (or yet to be found out as in experimental drugs
> undergoing testing).

You might find that. I just as soon subclass drug role with terms that
are more clearly defined that "connotation of beneficial". For example
an 'pain relief drug role' would be a drug role whose realization is
as a lessened transmission of pain signals, a decrease in origination
of signals from nociceptors, or a reduced perception of pain.

If you take the meaning of medication role to be in additional to drug
role, you could make a term using drug role and ogms:treatment.
medication role is a drug role that is realized as part of a
treatment.

> Re 'codeine', while the use of the word disease in the definition maybe too
> narrow, it nevertheless would be considered a 'medication' since it is used
> to treat a manifestation of a disease or a disorder.

I think the mention of disease or disorder is too strong. It misses
injury, prevention, etc.

Treatment: A processual entity whose completion is hypothesized (by a
healthcare provider) to alleviate the signs and symptoms associated
with a disorder

Is weaker (and more appropriate), though the definition of treatment
itself is too strong. We get preventative treatments such as fluoride
treatments, prenatal folate, vaccination.

A label that better conveys the sense of the definition would be
'treatment of a disorder'.

Also treatments try not only to alleviate the signs or symptoms, but
also to cure - i.e. remove the disorder, such as in autologous stem
cell transplant.

-Alan

-Alan

[Bill Hogan]

I still disagree that treatments are all-inclusive of prevention and
prophylaxis.

My medical training says that vaccinating kids is not treatment.

Bill

On Wed, Jun 20, 2012 at 1:26 AM, Alan Ruttenberg

[Alan Ruttenberg]

Use of language is, as usual, ambiguous. You can find, presenting
alternatives, the phrase "prevention or treatment" (45k google hits
for each of this and the reverse order) or you can find adjectival
modification suggesting subclassing as in "preventive treatment" (30k
hits + 65k hits for "prophylactic treatment".

Given that the adjectival use seems to be as common as the
alternatives use, my suggestion is to choose a label that is less
likely to be confused.

Finally, I think the way the statement is phrases exposes a common way
that things go wrong in ontological discussions. The form is:

<label> is not (or is) <label> where "is not", or "is" is taken to
mean (not) is_a or (not) instance_of depending on context.
This suggests that the label comes with a single definition (otherwise
how to parse the statement).

I conceive the order of development as follows:

Words or phrases are initial hooks into what the type of referents
are. We then work on definitions that give a more workable way to
delineate members of the class. Finally we go back and associate one
or more labels with the definition, sometimes constructing such labels
so they are unique within a scope (e.g. OBO foundry editor preferred
terms) or to capture the word members of a specific community use to
refer to the type as community specific label.

The best interpretation of what you can mean by "vaccinating kids is
not treatment" is that: within the community from which you received
your medical training, the label "treatment" is used to designate the
type defined also as: #1 "A processual entity whose completion is

hypothesized (by a healthcare provider) to alleviate the signs and

symptoms associate with a disorder" and "vaccinating kids" designates
a type that does not include #1 as there are no signs or symptoms or
disorder to hypothesize about.

(minding, of course, that we understand the sufficiently same
referents of the words defining #1 so as not to undermine the above
conclusion)

However, in the wider scope, as illustrated by google, such an
assessment would only be made by a portion of the people (say half),
while the opposite assessment might be made by the other half.
Therefore, in the interest of avoiding this confusion, I suggested
associating the label 'treatment of a disorder' with #1. I would have
"treatment" be associated with #1 as an alternative term as a
community specific label.

In a love/hate relationship with words,
-Alan

[Mathias Brochhausen]

This decision made me aware of that fact that I have some unclarity regarding the definition of the OBI term "drug role". I think we need some discussion about that.

Right now the definition of "drug role" reads

"a role borne by a molecular entity and is realized in a process of absorption by an organism  alters, or effects  (or is assumed to effect) a function(s) which inhere in an organism".

Bill Hogan once pointed out (correctly, in my opinion) that a single molecule of a drug does not have any effect at all and sometimes what creates the effect is a combination of molecular entities.

Furthermore, to me it sounds like what we are talking about in that definition (regardless of the granularity of the bearer) is a disposition (please see definition of "disposition") and not a role,since the physiological effects that are created are essential (s. definition of "role").

I propose the following:

create class "drug disposition"

which should use a amendment of the addition above (we need to look into the granularity issue)

define "drug role" by a social context (something like: role borne by a portion of a chemical entity that is inserted into an organism to alter and effect (a) function(s) inhering in that organism."

We could then have legal drug role which would be denied something like:

"role borne by a portion of a chemical entity that is approved by a legislative body for insertion into an organism to alex and effect ...)"

This class could have subclasses like FDA-approved drug-role in application ontologies.

Keep in mind that BFO allows to make a difference between bearing a role and playing a role. All bearers of a drug disposition can play the role of a drug regardless of the question whether they are intended to be used as drugs.

I am aware that this proposal will need amendment and additional work, but I wanted to put it out for discussion.

-Mathias

[Bill Hogan]

Alan,

I absolutely agree that figuring out the referent/definition first,
then applying labels second, is the right way to do things.

I agree with your proposal for "treatment" as an alternative term for
#1, and propose the label "medical intervention" or "health
intervention" (for the folks who dislike "medical" as being too
closely associated with physicians only) for #1.

But I agree the definitions are reasonably solid.

By the way, does OBI define "intervention" in the sense of an
"interventional trial"? We may want to coordinate there, somehow, if
so.

Also in a love/hate relationship with human language,

Bill

On Wed, Jun 20, 2012 at 10:13 AM, Alan Ruttenberg

[Alan Ruttenberg]

On Wed, Jun 20, 2012 at 11:19 AM, Mathias Brochhausen <mbroch...@googlemail.com> wrote:

This decision made me aware of that fact that I have some unclarity regarding the definition of the OBI term "drug role". I think we need some discussion about that.

Yes, I agree.

 

Right now the definition of "drug role" reads

"a role borne by a molecular entity and is realized in a process of absorption by an organism  alters, or effects  (or is assumed to effect) a function(s) which inhere in an organism".

Bill Hogan once pointed out (correctly, in my opinion) that a single molecule of a drug does not have any effect at all and sometimes what creates the effect is a combination of molecular entities.

Furthermore, to me it sounds like what we are talking about in that definition (regardless of the granularity of the bearer) is a disposition (please see definition of "disposition") and not a role,since the physiological effects that are created are essential (s. definition of "role").

The reason it is a role rather than disposition is in the "is assumed to effect". Some things with drug roles don't act as the purveyors suggest. The reason for the ascription to molecular entity is to make the link to chebi molecular entity clearer, and also to allow for homeopathic drugs (IIRC). Under some interpretations I think it would be more correct to say it inheres in some aggregate, but IIRC it was difficult to bound the nature of such, and with this interpretation of drug role we were able to express what was needed. Let's see what your proposal is..

 

I propose the following:

create class "drug disposition"

which should use a amendment of the addition above (we need to look into the granularity issue)

I think this could be in addition to drug role, but is there some combination of epistemological and non-monotinicity that this would give rise to? I'm thinking of a case where what we knew of the mechanism changed. Suppose we said something that was a gas of molecules with drug role, whose mechanism was based in binding to a receptor. Later we found instead that the real mechanism was in the reduction of partial pressure of oxygen during administration. Would we say that we were mistaken about the drug_role and fix it, or would we still label the gas as having molecules with drug role?

define "drug role" by a social context (something like: role borne by a portion of a chemical entity that is inserted into an organism to alter and effect (a) function(s) inhering in that organism."

I like the gist of this. The current drug role acknowledges intention only obliquely. Part of the reason that drugs are roles is that they are distinguished (I think) as involving some intentional aspect in their selection. We don't consider environmental arsenic to carry a drug role (I don't think) even though it is absorbed and alters bearers of function (many actually - too close to phosphorus for comfort). 

The other elements need to be examined. "insert" is probably too strong. You also expose people to drugs and their bodies can actively or passively absorb it. See below for a comment on "portion of chemical entity".

 

We could then have legal drug role which would be denied something like:

"role borne by a portion of a chemical entity that is approved by a legislative body for insertion into an organism to alex and effect ...)"

Yes. That was intended. 

This class could have subclasses like FDA-approved drug-role in application ontologies.

Keep in mind that BFO allows to make a difference between bearing a role and playing a role. All bearers of a drug disposition can play the role of a drug regardless of the question whether they are intended to be used as drugs.

What is this thing "playing a role" that you speak of?

 

I am aware that this proposal will need amendment and additional work, but I wanted to put it out for discussion.

Let's hear what others think. Portion of chemical entity is pretty much equivalent to material entity - that's the problem I was referring to earlier.

There's another issue with drug role as defined - it says "alters or effects". But sometimes it can actually bear a function the organism doesn't have (it might have - further subdivided as resulting from genetic disorder or loss or change of function bearers) or the function is novel, for example blocking the action of a accidental metabolite (e.g. poison antidotes).

In any case, the main thing I wanted to avoid, when the discussion came up at the OGMS session, was that OBI and OGMS went on their separate ways regarding this and similar terms. There are elements of the OGMS definition that are appealing as well as flaws and unappealing aspects, and the same is true of the OBI definition. Since this is a critical term, I think we need to think it carefully through (collaboratively!) to come up with the set of terms in the vicinity.

-Alan

[Mathias Brochhausen]

Alan: I absolutely agree that OBI and OGMS should not walk separate ways on that.

On Wed, Jun 20, 2012 at 11:21 AM, Alan Ruttenberg <alanrut...@gmail.com> wrote:

On Wed, Jun 20, 2012 at 11:19 AM, Mathias Brochhausen <mbroch...@googlemail.com> wrote:

This decision made me aware of that fact that I have some unclarity regarding the definition of the OBI term "drug role". I think we need some discussion about that.

Yes, I agree.

 

Right now the definition of "drug role" reads

"a role borne by a molecular entity and is realized in a process of absorption by an organism  alters, or effects  (or is assumed to effect) a function(s) which inhere in an organism".

Bill Hogan once pointed out (correctly, in my opinion) that a single molecule of a drug does not have any effect at all and sometimes what creates the effect is a combination of molecular entities.

Furthermore, to me it sounds like what we are talking about in that definition (regardless of the granularity of the bearer) is a disposition (please see definition of "disposition") and not a role,since the physiological effects that are created are essential (s. definition of "role").

The reason it is a role rather than disposition is in the "is assumed to effect". Some things with drug roles don't act as the purveyors suggest. The reason for the ascription to molecular entity is to make the link to chebi molecular entity clearer, and also to allow for homeopathic drugs (IIRC). Under some interpretations I think it would be more correct to say it inheres in some aggregate, but IIRC it was difficult to bound the nature of such, and with this interpretation of drug role we were able to express what was needed. Let's see what your proposal is..

I think that we can take care of the assumed effect which we can keep in the definition of drug role (via the intention), but not have it in the "drug disposition", since there is no place for assumptions there. All we need to do is not to link drug disposition as closely to drug role as I insinuated in my previous mail. So, people can use something as a drug, regardless of the fact that it might not (really) have the corresponding disposition - they just assume it has.

 

I propose the following:

create class "drug disposition"

which should use a amendment of the addition above (we need to look into the granularity issue)

I think this could be in addition to drug role, but is there some combination of epistemological and non-monotinicity that this would give rise to? I'm thinking of a case where what we knew of the mechanism changed. Suppose we said something that was a gas of molecules with drug role, whose mechanism was based in binding to a receptor. Later we found instead that the real mechanism was in the reduction of partial pressure of oxygen during administration. Would we say that we were mistaken about the drug_role and fix it, or would we still label the gas as having molecules with drug role?

define "drug role" by a social context (something like: role borne by a portion of a chemical entity that is inserted into an organism to alter and effect (a) function(s) inhering in that organism."

I like the gist of this. The current drug role acknowledges intention only obliquely. Part of the reason that drugs are roles is that they are distinguished (I think) as involving some intentional aspect in their selection. We don't consider environmental arsenic to carry a drug role (I don't think) even though it is absorbed and alters bearers of function (many actually - too close to phosphorus for comfort). 

The other elements need to be examined. "insert" is probably too strong. You also expose people to drugs and their bodies can actively or passively absorb it. See below for a comment on "portion of chemical entity".

Sure, I agree on the insert stuff. "expose" makes more sense. 

 

We could then have legal drug role which would be denied something like:

"role borne by a portion of a chemical entity that is approved by a legislative body for insertion into an organism to alex and effect ...)"

Yes. That was intended. 

This class could have subclasses like FDA-approved drug-role in application ontologies.

Keep in mind that BFO allows to make a difference between bearing a role and playing a role. All bearers of a drug disposition can play the role of a drug regardless of the question whether they are intended to be used as drugs.

What is this thing "playing a role" that you speak of?

So, I am not the bearer of a pilot role. But in an emergency situation where the pilot of an aircraft p[asses out and I grab the steering device, I actually play the role of a pilot without being a bearer. This is elaborated in the BFO2 reference, but I think that it was a distinction always made my (at least) Barry . 

 

I am aware that this proposal will need amendment and additional work, but I wanted to put it out for discussion.

Let's hear what others think. Portion of chemical entity is pretty much equivalent to material entity - that's the problem I was referring to earlier.

I agree that this part of the definition of "drug disposition" is the hard not, beyond the fact that I still advocate keeping drug role and drug deposition separate. 

For a smallish project we run here on drug ontology, we will assign drug roles to actual pills and solutions as a whole since that is what gets approved by the FDA (so, actually I think we use a subtype of legal drug role)

Mathias