[Obi-devel] objective discussion

[Bjoern Peters]

All,

The discussion of 'objectives', 'plans', 'planned process' and the
relations between them have been held up for quite a while because of
dependencies between the denrie branch, plans + planned process branch,
role branch and bfo. Terms like these are clearly where our branch
structure shows weaknesses, as every and no branch is responsible for
the final conclusion. The dev-call is where we should solve this, which
is why Melanie suggested to put the 'objective' term on the agenda for
Wednesdays. Read on if you care about that agenda item, and please join
the call or email if you have any input that needs to be considered.

* We have had quite a lot of discussions on objectives in the plan +
planned process branch, and I had some email back and forth with the
denrie branch. One key point that came up over and over was if
objectives themselves are realizable. I followed up with the bfo list on
this, as it is clearly not obi territory. Check:
http://groups.google.com/group/bfo-discuss/t/a22b211aaab8577f

* I had already modified the obi svn owl files to implement distinctions
between investigations and protocol applications under 'process', which
relate to the objective discussion. After doing that I realized that it
isn't a good idea to modify the high level hierarchy used by everyone
for a debatable solution to one issue. At the same time, I have
continuously run into problems communicating what I want to implement
without putting it into owl. Therefore, I have now created a svn:branch
of the .owl files that incorporate the changes to the information object
hierarchy I am proposing. Please look at that, by checking out:
https://obi.svn.sourceforge.net/svnroot/obi/branches/bpbranch.

I hope this will be a useful starting point for Wednesdays call.

- Bjoern

Melanie Courtot wrote:
> Hi Bjoern,
>
> I tentatively added objective to the agenda for next dev call - cf
> https://wiki.cbil.upenn.edu/obiwiki/index.php/ConferenceCallAgenda#May_2008
>
> It might be a lot on your plate though as you are already dealing with
> function (on the same agenda)
>
> I don't know what is the status of objective implementation/discussion?
> Currently, objective is under non realizable entity.
>
> I don't think this is definitive though?
> Is it ok to add subclasses there as needed?
>
> To take a DT example (hence the cc to James ), if I want to add an
> objective called "normalization".
> Can I add as subclass of methodology_testing_objective ->
> protocol_testing_objective? Shall I instead submit to DENRIE which
> will deal with that? Should I submit to PlanAndPlannedProcess instead?
>
> There are several objective related tracker items as well:
> - objective model
> https://sourceforge.net/tracker/index.php?func=detail&aid=1887174&group_id=177891&atid=886178
>
> - objective terms from DT
> https://sourceforge.net/tracker/index.php?func=detail&aid=1888956&group_id=177891&atid=886178
>
> - objective vs objective-neutral planned process
> https://sourceforge.net/tracker/index.php?func=detail&aid=1889390&group_id=177891&atid=886178
>
>
> Let me know if you think we can deal with that during next call or not.
>
> Thanks,
> Melanie
>
>

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Obi-devel mailing list
Obi-...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/obi-devel

[Cristian Cocos]

> Melanie suggested to put the 'objective' term on the agenda for
> Wednesdays. Read on if you care about that agenda item, and please join
> the call or email if you have any input that needs to be considered.

If I may follow up on my previous interventions linking "objective" to
"state of affairs," let me suggest that the way that seems to me to be the
most ontologically kosher in dealing with the former is to *not* turn it
into a class, but capture the idea of "objective" exclusively via the
objectiveOf/hasObjective relation:

StateOfAffairs objectiveOf Action

Action hasObjective StateOfAffairs

The trouble, however, is that the relata to be linked via these relations
seem to be falling outside of the scope of OBI; I am, as a matter of fact,
slightly skeptical about whether BFO in its present guise can deal with them
properly. While StateOfAffairs may be seen, by some stretch, as a
span:ProcessBoundary, I honestly don't know where to place Action in the BFO
scheme. Process? Child of Process? None of these looks to me to be good
enough. Besides, "objective" sometimes connotes potentiality, as opposed to
actuality: we not only speak of *actual* states of affairs that obtained as
a result of past actions, but *hypothetical* states of affairs that might or
might not obtain as a result of future actions.

Substituting "Plan" for "Action" also doesn't look to me to be a good move
as plans are more akin to abstract ("informational") entities, while actions
to processes.

CC

[Daniel Schober]

Hi Christian,
I tried to use a section from our use case at http://docs.google.com/Doc?docid=dg9kgrmp_1hjv9jnf6&hl=en

As an example we can look at a rather low level objective from our use case (2.3.3.1  Preparation of Density Gradients). It might be difficult to decide at what level we need to model an objective (only for top level abstractions or for each process steps?):

1.	Prepare [state of affairs] Percoll solutions of the following densities (1.064, 1.075, 1.081, and 1.087 g/ml) [Action/Prosess] by diluting Percoll medium with 10× PBS and distilled water according to the manufacturer’s one-step procedure.

So each small protocoll part can have its own objective:
Here it seems the objective/state of affairs is to have 4 Percoll solutions of defined density.
The Action/Process to archieve this objective is the dilution of a Medium with buffers and water according to PA/manufacturers_one-step_procedure .

 I would Represent:

 State of affairs: (I see this as a list of assertions or facts in the knowledgebase)

Fact 1: Percoll solution has_density 1.064 g/ml
...
Fact 4: Percoll solution has_density 1.087 g/ml

 Action/Process: (Where is the problem you have with Processes as Actions?)

Process/manufacturers_one-step_procedure: (modeled as composite)
    Percoll medium dilute_with solution/10xPBS
    Percoll medium dilute_with destilled water
    ...

I see 'states of affairs' as a set of assertions/facts within the knowledgebase. So, the range and domain of the assertion you propose ( Action hasObjective StateOfAffairs ) is here broken down into many assertions. Is that right? Further..., is your notion of 'state of affairs' the same as the one from Wittgenstein? From  http://www.formalontology.it/wittgensteinl.htm

"The ontological part of the Tractatus occupying its first few pages consists of 49 theses: 

1-2063 and concerns 65 notions —from "the world" and "what is the case" in 1 to the "independent" in 2.061 — introduced with the frequency vrying from 21 uses of one notion (object — Gegenstand and state of affairs—Sachverhalt) to notions mentioned only once. From the frequency point of view the ontology of the Tractatus is the ontology of objects and states of affairs, but understanding it as the ontology of objects, states of affairs and facts is more common and reasonable.

A very brief account of the Tractatus ontology is as follows: The world is the totality of facts, facts are constituted by states of affairs consisting of objects standing in relations to each other. Objects are simples, the rest consists of complex items (states of affairs, facts, situations, the world): What is complex has a structure, i.e. the way objects hang together in the item and the stuff (or substance), i.e. a collection of objects included in the item: The object is the item which is constant; fixed and necessary, whereas the configuration of objects (complex item!) is the item which is changeable and contingent. Which configuration is possible is determined by internal (essential) properties of objects entering into a given configuration, by their nature. Let me recall 2.012: In logic nothing is accidental, if a thing can occur in a state of affairs, the possibility of the state of affairs must be written into the thing itself." p. 224-225"

Am I right to correlate Wittgenstein terminology to our OWL DL speak in the following way ?:
state of affairs= set of assertions/triples/facts
Object =atomic class
complex item = composite/defined class (but also set of assertions?)

I would like to know because we used the term 'states of affairs' in OBI before, e.g. see my obi-dev mail from  02/11/2007 13:46
( http://groups.google.com/group/obi-developer/msg/e1c08ba4bcc4773c?hl=en& )

By the way, I think we should have such discussions here at the list OR at the SF Tracker (see
https://sourceforge.net/tracker/index.php?func=detail&aid=1889390&group_id=177891&atid=886178 ), but not at both as we do it currently.

Cheers,
    Daniel Schober

-- 
__________________________________________________________________________________________

Dr. Daniel Schober

NET Project - Ontologist

The European Bioinformatics Institute   email:  sch...@ebi.ac.uk
EMBL Outstation - Hinxton               direct: +44 (0)1223 494410
Wellcome Trust Genome Campus            fax: +44 (0)1223 494 468
Cambridge CB10 1SD, UK                 	Room: A3-141 (extension building)

Project page: www.ebi.ac.uk/net-project

Personal page:    http://www.ebi.ac.uk/Information/Staff/person_maint.php?s_person_id=734
Former home page: http://www.bioinf.mdc-berlin.de/%7Eschober/

[Alan Ruttenberg]

FYA, here's recent cull from a table in the Coriell catalog that we
are starting at looking to represent in the neurocommons.
It's called "Research Intent", and it sounds like it is in the space
of objectives:

Perform functional studies
Positive/negative controls
SNP discovery/genotyping/haplotyping
Probe development
Map genes
Other
Conservation
Evolutionary studies
Identify novel genes
Characterize genes and mutations
Study gene expression
Determination of ancestral alleles
Study molecular phylogenetics
Conduct proteomic studies
Haplotype mapping project
Presence/absence of genes across species
Identify regulatory elements
Natural selection
Compare repetitive elements
Sequence portions of the genome
Use as pos or neg control for genetic testing
Use as pos or neg control for assay development
Evolutionary analysis of human variation
DNA methylation
Education
Use data as a control for other studies
Add additional samples to an existing control population
Part of a Whole Genome Association (WGAS)
For comparison of genotyping to my dataset from other cohorts
Search for novel variants in genes
Search for samples with specific genotypes/copy number variants
Search for SNP/copy number variants structure within populations
Identify specific samples for use as reference materials
For proficiency testing
Whole genome sequencing

-Alan

[Fostel, Jennifer (NIH/NIEHS) [C]]

thanks, Bjorn and Alan. the discussion is interesting and helpful.
from the bfo discussion (at the moment), the objective will describe a desired terminal boundary for reality after a process. these seem to fit that definition, and align with intuition.

do we need to create a hierarchy for objectives? or would this follow from the plan / protocol application hierarchy?

...jennifer

[Philippe]

Thanks Alan for this

Interestingly, most of them are process or protocol applications.
(lots of verbs or verbal phrases)

Philippe

[Fostel, Jennifer (NIH/NIEHS) [C]]

taking the example of map gene, it could be a protocol application map gene, and also the objective of a PA if we rephrased it as mapped gene. the objective described the reality of having mapped the gene.

-----Original Message-----
From: Philippe [mailto:ro...@ebi.ac.uk]
Sent: Wed 5/28/2008 11:10 AM
Cc: 'OBI Developers'
Subject: Re: [Obi-devel] objective discussion