[Obi-protocol-application-branch] Protocol Application review: consistency/style of restriction
Philippe Rocca-Serra
Following Alan's note on obsolete relation, I am going over the protocol
application branch and checking restrictions:
I have so far spotted a few errors such as
DNA sequence feature detection:
has_specified_input some 'double-stranded DNA' or
DNA sequence variation detection:
has_specified_input some 'ribonucleic acids' **
but the more important is the following:
*animal feeding* has currently the following restriction: realizes some
'feed role' , which should probably be replaced by:
realizes some ('feed role' and (inheres_in some 'material_entity'))....
or should it be
has_specified_input some (material_entity and (realizes some 'feed role'))
Classifier likes either but I guess it is more than a simple question of style. Bjoern had pointed to frequent abuse of 'has_specified_input' relation.
I have found myself confronted with this issue with several classes already (e.g. elution...)
s already (e.g. elution with a restriction such as has_specified_input some ('{molecular entities, molecular entity}' and (has_role some 'solvent role')))
Cheers
--
Philippe Rocca-Serra, PhD
Technical Coordinator
www.ebi.ac.uk/net-project
The European Bioinformatics Institute email: ro...@ebi.ac.uk
EMBL Outstation - Hinxton direct: +44 (0)1223 492 553
Wellcome Trust Genome Campus fax: +44 (0)1223 492 620
Cambridge CB10 1SD, UK room: A3-141
--
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Obi-protocol-application-branch mailing list
Obi-protocol-ap...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
Bjoern Peters
> *animal feeding* has currently the following restriction: realizes some
> 'feed role' , which should probably be replaced by:
>
> realizes some ('feed role' and (inheres_in some 'material_entity'))....
>
>
Not sure what the .. is. If you want to connect it to
'is_specified_participant', that won't work.
> or should it be
>
> has_specified_input some (material_entity and (realizes some 'feed role'))
>
>
This does not work, as 'realizes' needs a process on the left hand
side, not a material entity. What does work is 'has_role'.
> Classifier likes either but I guess it is more than a simple question of style. Bjoern had pointed to frequent abuse of 'has_specified_input' relation.
> I have found myself confronted with this issue with several classes already (e.g. elution...)
> s already (e.g. elution with a restriction such as has_specified_input some ('{molecular entities, molecular entity}' and (has_role some 'solvent role')))
>
>
I agree that we need to have one consistent way of modeling this. I
don't believe we can logically exactly say that an entity is a specified
participant of a process and the role it realizes.
>
--
Bjoern Peters
Assistant Member
La Jolla Institute for Allergy and Immunology
9420 Athena Circle
La Jolla, CA 92037, USA
Tel: 858/752-6914
Fax: 858/752-6987
http://www.liai.org/pages/faculty-peters
Frank Gibson
I believe the above is the best we can do
> *animal feeding* has currently the following restriction: realizes some
> 'feed role' , which should probably be replaced by:
>
Ok, so the class as it is is wrong, it probably shoudl be
animal feeding
has_specified_input some organism
has_specified_input some (material has_role feed_role)
has_specified_ouput some fatter_organism
The output here has to indicate that the organims has been fed.
Frank
--
Frank Gibson, PhD
http://peanutbutter.wordpress.com/
Philippe Rocca-Serra
+1
Frank, I believe we still need to state process 'animal feed' realizes
feed_role
since the has_specified_input relation does not imply anything about the
realization of a role born by the continuant being specified
if this assertion is wrong then the has_specified_input relation needs
to be clarified and an example supplied
Cheers
P
> <mailto:Obi-protocol-ap...@lists.sourceforge.net>
> https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
>
>
>
>
> --
> Frank Gibson, PhD
> http://peanutbutter.wordpress.com/
--
Philippe Rocca-Serra, PhD
Technical Coordinator
www.ebi.ac.uk/net-project
The European Bioinformatics Institute email: ro...@ebi.ac.uk
EMBL Outstation - Hinxton direct: +44 (0)1223 492 553
Wellcome Trust Genome Campus fax: +44 (0)1223 492 620
Cambridge CB10 1SD, UK room: A3-141
--
Frank Gibson
Frank, Bjoern
+1
Frank, I believe we still need to state process 'animal feed' realizes feed_role
I think you need to have this relation on the role so the. feed_role is_realized_by some animal_feeding_process
Frank
James Malone
priority 9 and we can add it to the call monday and assign the task.
I'm not sure how critical it is, if anyone thinks so could you do that
please (if it given 9 I will add it to agenda). If it's not critical we
should probably still add it so it's not lost.
Cheers,
James
Frank Gibson wrote:
>
>
> On Thu, Jul 30, 2009 at 4:11 PM, Philippe Rocca-Serra <ro...@ebi.ac.uk
> <mailto:ro...@ebi.ac.uk>> wrote:
>
> Frank, Bjoern
>
> +1
>
> Frank, I believe we still need to state process 'animal feed'
> realizes feed_role
>
>
> I think you need to have this relation on the role so the. feed_role
> is_realized_by some animal_feeding_process
>
> Frank
>
>
>
>
>
>
> since the has_specified_input relation does not imply anything
> about the realization of a role born by the continuant being specified
>
> if this assertion is wrong then the has_specified_input relation
> needs to be clarified and an example supplied
>
> Cheers
>
> P
>
> Frank Gibson wrote:
>
>
>
> On Thu, Jul 30, 2009 at 3:46 PM, Bjoern Peters
> <bpe...@liai.org <mailto:bpe...@liai.org>
> <mailto:Obi-protocol-ap...@lists.sourceforge.net
> <mailto:Obi-protocol-ap...@lists.sourceforge.net>>
>
>
> https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
>
>
>
>
> --
> Frank Gibson, PhD
> http://peanutbutter.wordpress.com/
>
>
>
> --
> Philippe Rocca-Serra, PhD
>
> Technical Coordinator
> www.ebi.ac.uk/net-project <http://www.ebi.ac.uk/net-project>
>
> The European Bioinformatics Institute email: ro...@ebi.ac.uk
> <mailto:ro...@ebi.ac.uk>
> EMBL Outstation - Hinxton direct: +44 (0)1223 492 553
> Wellcome Trust Genome Campus fax: +44 (0)1223 492 620
> Cambridge CB10 1SD, UK room: A3-141
> --
>
>
>
>
> --
> Frank Gibson, PhD
> http://peanutbutter.wordpress.com/
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> ------------------------------------------------------------------------
>
> _______________________________________________
> Obi-devel mailing list
> Obi-...@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/obi-devel
>
--
European Bioinformatics Institute,
Wellcome Trust Genome Campus,
Hinxton,
Cambridge, CB10 1SD,
United Kingdom
Tel: + 44 (0) 1223 494 676
Fax: + 44 (0) 1223 492 468
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Obi-protocol-application-branch mailing list
Obi-protocol-ap...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
Frank Gibson
If this is important it should be added to the tracker and assigned a priority 9 and we can add it to the call monday and assign the task. I'm not sure how critical it is, if anyone thinks so could you do that please (if it given 9 I will add it to agenda). If it's not critical we should probably still add it so it's not lost.
This is the way we define roles, feed_role is missing the realised in feeding_process restriction. You dont need to state it again in the process, we can infer it.
Frank
Bjoern Peters
to logically exactly say that an entity is a specified participant of a process AND the role it realizes in that process..
Frank, Bjoern
+1
Frank
Cheers
P
Frank Gibson wrote:
Frank
https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
--
Philippe Rocca-Serra, PhD
Technical Coordinator
www.ebi.ac.uk/net-project
https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
Frank Gibson
Frank: that is not the point. The individual statements are trivial. The problem is
to logically exactly say that an entity is a specified participant of a process AND the role it realizes in that process..
You cant - and it should not be something you want to do. The material doesnt realise a role, it bears it (has_role), for the duration of the process. Becasue is bears the role and the process complete the role has been realized by the process, which includes the participants. I admit the BFO defintion of role is not very helpful - there is a surprise...
You can ask what role is it bearing for the duration of the process and infer it was realised. You cant ask your statement above.
Frank
Bjoern Peters
role is realized in that process, at least according to Barry's
definitions. Anyway, I was ignoring role / function differences in this
discussion, in anticipation of replacing it with the solution presented
by Phil at the workshop. 'feed role' here is a good example why we need
that, as many of the things being fed to lab animals were manufactured
for that purpose, and therefore would have a 'feed function' from the
start of their existence. I forgot if we agreed on how to call the
parent entities of function / role , so I will rewrite using 'capacity'
animal feeding has_specified_participant some material entity and
bearer_of some feed capacity
AND
animal feeding realizes feed capacity borne by some material entity
What we can't say in OWL is that the same material entity is referred
to. Even if this may not be a problem for role if you believe every time
a material has a role it also realizes that role, it is a problem for
functions. A process realizing the plann to use a bag of manufactured
animal feed as a door stopper would fulfill the first statement above.
The second statement avoids that, but we don't get a reference to the plan.
- Bjoern
> <mailto:bpe...@liai.org> <mailto: bpe...@liai.org
> <mailto: Obi-protocol-ap...@lists.sourceforge.net
> <mailto:Obi-protocol-ap...@lists.sourceforge.net> >
>
> https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
>
>
>
>
> --
> Frank Gibson, PhD
> http://peanutbutter.wordpress.com/
>
>
>
>
>
> --
> Philippe Rocca-Serra, PhD
>
> Technical Coordinator
> www.ebi.ac.uk/net-project <http://www.ebi.ac.uk/net-project>
>
> The European Bioinformatics Institute email: ro...@ebi.ac.uk
> <mailto:ro...@ebi.ac.uk>
> EMBL Outstation - Hinxton direct: +44 (0)1223 492 553
> Wellcome Trust Genome Campus fax: +44 (0)1223 492 620
> Cambridge CB10 1SD, UK room: A3-141
> --
>
>
>
>
> --
> Frank Gibson, PhD
> http://peanutbutter.wordpress.com/
>
>
>
>
> --
> Frank Gibson, PhD
> http://peanutbutter.wordpress.com/
--
Bjoern Peters
Assistant Member
La Jolla Institute for Allergy and Immunology
9420 Athena Circle
La Jolla, CA 92037, USA
Tel: 858/752-6914
Fax: 858/752-6987
http://www.liai.org/pages/faculty-peters
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Obi-protocol-application-branch mailing list
Obi-protocol-ap...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
Philippe Rocca-Serra
We need both statement to fully qualify the process we are describing.
If we all agree on this, (please chip in if you disagree) then we can
start correcting /adjusting the restrictions of a number of processes (I
have already identified 12 classes and was not even half way through ).
Cheers
P
--
Philippe Rocca-Serra, PhD
Technical Coordinator
www.ebi.ac.uk/net-project
Bjoern Peters
Just now I am thinking:
How about a relation planned_realization_of' (crappy label for now)=def: a relationship between a planned process and a realizable dependent continuant for which the planned process realizes the realizable dependent continuant, and for which the plan specification associated with the planned process specifies the realization of the dependent continuant.
Then we can say:
animal feeding planned_realization_of some feed capacity
- Bjoern
Melanie Courtot
I think I don't get the difference between the proposed
planned_realization_of and realizes.
definition of realizes: "Relation between a process and a function,
where the unfolding of the process requires the execution of the
function. Class level: P realizes F iff: given any p that instantiates
P, there exists some f, t such that f instantiates F at t and p
*realizes* f. Here, *realizes* is the primitive instance level
relation [GOC:cjm]"
This probably needs to be updated to say function and roles, which
would match the range realizable_entity.
By using this relation with planned processes we would get the plan
specification for free, as planned_process is defined as realization
of concretization of a plan.
Or were you suggesting to create a new relation to restrain the
current domain of realizes (i.e. process) to only planned processes?
Melanie
ps: I think Phil's proposal was to assert directly under realizable
entities.
> Obi-devel mailing list
> Obi-...@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/obi-devel
---
Mélanie Courtot
TFL- BCCRC
675 West 10th Avenue
Vancouver, BC
V5Z 1L3, Canada
Frank Gibson
I think I don't get the difference between the proposed
planned_realization_of and realizes.
I would agree with you here Melanie. We seem to be on slightly differenty wavelengths here Bjoern. We can only build against the version of BFO that we have at the minute - to do anything else is a recipie for disaster. These have specfic defintions for what a realizable_entiy is and how is is realized. You seem to be talking about advanced versions of BFO that inlcude capacity - I have no idea what capacity is. What I describe above is how to model the example currently, and what we can not do currently. I think we need these boundaries to get version 1 and the paper out.
definition of realizes: "Relation between a process and a function,
where the unfolding of the process requires the execution of the
function. Class level: P realizes F iff: given any p that instantiates
P, there exists some f, t such that f instantiates F at t and p
*realizes* f. Here, *realizes* is the primitive instance level
relation [GOC:cjm]"
This probably needs to be updated to say function and roles, which
would match the range realizable_entity.
Or all realizable entities? As Melanie points out here, some thing is not correct here, either the definition or the domain and range.
By using this relation with planned processes we would get the plan
specification for free, as planned_process is defined as realization
of concretization of a plan.
Which is not the current textual definition of the realizes relation.
Or were you suggesting to create a new relation to restrain the
current domain of realizes (i.e. process) to only planned processes?
I think we have overloaded the realizes relation here, we need some clarification on what is actually is for, improving the textual def and the domain and range. Keeping in mind we already have the inheres_in relation which links realizable entities with their bearers - maybe the realizes relation should have the domain realizable_entity_bearer (some material bearing a realizable entity) and the range process. Or something else completely different........
ps: I think Phil's proposal was to assert directly under realizable
entities.
It was
Frank
Bjoern Peters
>
> On Jul 30, 2009, at 8:01 AM, Frank Gibson wrote:
>
>> On Thu, Jul 30, 2009 at 3:46 PM, Bjoern Peters <bpe...@liai.org> wrote:
>>
>> > *animal feeding* has currently the following restriction: realizes
>> some
>> > 'feed role' , which should probably be replaced by:
>> >
>> I believe the above is the best we can do
>>
>> Ok, so the class as it is is wrong, it probably shoudl be
>>
>> animal feeding
>> has_specified_input some organism
>> has_specified_input some (material has_role feed_role)
>> has_specified_ouput some fatter_organism
>>
>> The output here has to indicate that the organims has been fed.
>
> This is fine if the above is intended as necessary conditions, but if
> the goal is to produce N+S conditions then this is problematic. You're
> not really stating that the input organisms have been fed. The input
> organisms may have been ground up and fed to an entirely different
> organism.
>
> Jumping in halfway here, apologies if this has already been discussed
>
I raised this in an earlier thread: We are implicitly assuming a
'derived from' like relationship between the specified inputs and
specified outputs of a process. We have not been able to either state
this effectively in OWL or figured out how exactly to define that
relation. At the same time, we also did not run into practical problems
with this yet.
Bjoern
>> Obi-devel mailing list
>> Obi-...@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/obi-devel
>
Bjoern Peters
role / function is specifically part of the plan. This is identical to
how we distinguish participation and specified participation.
https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
Bjoern Peters
order to identify the 'realizable entity' parent of 'animal feed role'
and 'animal feed function'.
If you want to get anal about the inconsistencies in the RO definition
of realizes and our extension of it to other realizable entities (which
we have done for years), then I do give up now.
> >>> <mailto:obi-...@lists.sourceforge.net
> <mailto:obi-...@lists.sourceforge.net>>>, "Protocol App Branch"
> >>> <obi-protocol-ap...@lists.sourceforge.net
> <mailto:obi-protocol-ap...@lists.sourceforge.net>
> >>>
> <mailto:obi-protocol-ap...@lists.sourceforge.net
> <mailto:obi-protocol-ap...@lists.sourceforge.net>>>
> >>> Sent: Thursday, July 30, 2009 8:29:48 AM GMT -08:00 US/Canada
> >>> Pacific
> >>> Subject: Re: [Obi-protocol-application-branch] Protocol
> >>> Application review: consistency/style of restriction
> >>>
> >>>
> >>> On Thu, Jul 30, 2009 at 4:11 PM, Philippe Rocca-Serra <
> >>> ro...@ebi.ac.uk <mailto:ro...@ebi.ac.uk>
> <mailto:ro...@ebi.ac.uk <mailto:ro...@ebi.ac.uk>> > wrote:
> >>>
> >>>
> >>> Frank, Bjoern
> >>>
> >>> +1
> >>>
> >>> Frank, I believe we still need to state process 'animal feed'
> >>> realizes feed_role
> >>>
> >>> I think you need to have this relation on the role so the.
> >>> feed_role is_realized_by some animal_feeding_process
> >>>
> >>> Frank
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> since the has_specified_input relation does not imply anything
> >>> about the realization of a role born by the continuant being
> >>> specified
> >>>
> >>> if this assertion is wrong then the has_specified_input
> relation
> >>> needs to be clarified and an example supplied
> >>>
> >>> Cheers
> >>>
> >>> P
> >>>
> >>> Frank Gibson wrote:
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>> On Thu, Jul 30, 2009 at 3:46 PM, Bjoern Peters <
> bpe...@liai.org <mailto:bpe...@liai.org>
> >>> <mailto:bpe...@liai.org <mailto:bpe...@liai.org>>
> <mailto: bpe...@liai.org <mailto:bpe...@liai.org>
> <http://www.ebi.ac.uk/net-project> <http://www.ebi.ac.uk/net-project>
> >>>
> >>> The European Bioinformatics Institute email:
> ro...@ebi.ac.uk <mailto:ro...@ebi.ac.uk>
> >>> <mailto:ro...@ebi.ac.uk <mailto:ro...@ebi.ac.uk>>
> >>> EMBL Outstation - Hinxton direct: +44 (0)1223 492 553
> >>> Wellcome Trust Genome Campus fax: +44 (0)1223 492 620
> >>> Cambridge CB10 1SD, UK room: A3-141
> >>> --
> >>>
> >>>
> >>>
> >>>
> >>> --
> >>> Frank Gibson, PhD
> >>> http://peanutbutter.wordpress.com/
> >>>
> >>>
> >>>
> >>>
> >>> --
> >>> Frank Gibson, PhD
> >>> http://peanutbutter.wordpress.com/
> >>
> >>
> >
> >
> > --
> > Philippe Rocca-Serra, PhD
> >
> > Technical Coordinator
> > www.ebi.ac.uk/net-project <http://www.ebi.ac.uk/net-project>
> >
> > The European Bioinformatics Institute email: ro...@ebi.ac.uk
> <mailto:ro...@ebi.ac.uk>
> > EMBL Outstation - Hinxton direct: +44 (0)1223 492 553
> > Wellcome Trust Genome Campus fax: +44 (0)1223 492 620
> > Cambridge CB10 1SD, UK room: A3-141
> > --
> >
> >
> >
> ------------------------------------------------------------------------------
> > Let Crystal Reports handle the reporting - Free Crystal Reports 2008
> > 30-Day
> > trial. Simplify your report design, integration and deployment - and
> > focus on
> > what you do best, core application coding. Discover what's new with
> > Crystal Reports now. http://p.sf.net/sfu/bobj-july
> > _______________________________________________
> > Obi-devel mailing list
> > Obi-...@lists.sourceforge.net
> <mailto:Obi-...@lists.sourceforge.net>
> > https://lists.sourceforge.net/lists/listinfo/obi-devel
>
> ---
> Mélanie Courtot
> TFL- BCCRC
> 675 West 10th Avenue
> Vancouver, BC
> V5Z 1L3, Canada
>
>
>
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports
> 2008 30-Day
> trial. Simplify your report design, integration and deployment -
> and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Obi-protocol-application-branch mailing list
> Obi-protocol-ap...@lists.sourceforge.net
> <mailto:Obi-protocol-ap...@lists.sourceforge.net>
> https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
>
>
>
>
> --
> Frank Gibson, PhD
> http://peanutbutter.wordpress.com/
--
Bjoern Peters
Assistant Member
La Jolla Institute for Allergy and Immunology
9420 Athena Circle
La Jolla, CA 92037, USA
Tel: 858/752-6914
Fax: 858/752-6987
http://www.liai.org/pages/faculty-peters
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Obi-protocol-application-branch mailing list
Obi-protocol-ap...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
Frank Gibson
As Melanie points out as well, the definition, the domain and range and the use of the realizes relation do not match. This needs to be addressed for the 1.0 release. I have added this issue to the tracker.
Frank