Conference calls Relations 2008 notes

09-12-2998
Call cancelled

November the 25th

 * Please comment on agenda items add any new items to the bottom of the list

[edit] Agenda Items

1/Please add your names as editors on the Relations home page

2/ Outstanding actions to be closed


 * submitted by the investigation branch

2.1/


 * process "executes" realizableenduringentity
 * study "executes" study_design

1. A similar relation will be added to the OBO Relation Ontology very soon.


 * studydesign "isexecuted_by" study

1. This is not true. There are many study designs which are never executed. Remember the rule (from http://genomebiology.com/2005/6/5/R46) is that A R B means: every instance of A stands in R to some instance of B

Action: closed - executed is a synonym of realizes

2.2/


 * studydesign "hasfactor" owl:Thing

1. A good biomedical ontology should be independent of any given computer realization; this is because computer programming languages and ontology representation languages change, and it should not be necessary to revamp the ontology every time such changes are made. 2. 'owl:Thing' is an artifact of a certain computer language. It is out of place in a biomedical ontology. There were study designs for many decades before there were entities properly designated by 'owl:Thing', and there will likely be study designs for many millennia long after OWL has fallen into disrepair. 3. Biologists should be able to use a well-structured biomedical ontology without needing to learn OWL (or any other computational idiom).


 * owl:Thing "isfactorin" study_design

Action: closed. There is an open tracker item on the role branch for experimental factor when it is determined what a factor is, the request for a relation can be re-submitted.

2.3/ [edit] Biomaterial branch
 * relationships between biomaterials to indicate a change has taken place, e.g. from starting material to the material after some process has been performed on the biomaterial
 * indicate that the characteristic of the biomaterial is a factor of the investigation

Action: closed. This is vague. A request has been made to the biomaterial branch to see if this is still relevant and if so to submit a tracker item for it.

2.4/ [edit] Protocol Application branch

1. Domain: protocol_application 2. Range: has_role input_role 3. Inverse: is_input_for
 * has_input - subclass of has_participant


 * has_output - subclass of has_participant

1. Domain: protocol_application 2. Range: has_role output_role 3. Inverse: is_output_for


 * utilizes_reagent - subclass of has_participant

1. Domain: Process 2. Range: continuant and has_role reagent_role 3. Inverse: is_reagent_in


 * utilizes_instrument - subclass of has_participant

1. Domain: Process 2. Range: instrument 3. Inverse: is_instrument_for


 * is_proxy_for

1. Domain: information_content_entity 2. Range: quality 3. Inverse: is_represented_by 4. Example: The A260/A280 of a nucleic acid sample is_proxy_for its level of purity.


 * has_role

1. Domain: Thing 2. Range: role

Action: closed: All these relations have been added, although there domain and ranges are slightly different.

3/ Definition of has_specfied_input/output as per tracker discussion.

4/Definition of is_concretization_of The following definition has been proposed via the issue tracker

is_concretization_of relates a generically dependent continuant to a specifically dependent continuant. A generically dependent continuant may inhere in more than one entity. It does so by virtue of the fact that there is, for each entity that it inheres, a specifically dependent specifically dependent. For instance, consider a story, which is an information artifact that inheres in some number of books. Each book bears some quality that carries the story. The relation between this quality and the generically dependent continuant is that the former is the concretization of the latter.
 * concretization* of the generically dependent continuant that is

Definition source: Alan channeling Barry.

5/The definition of inputs/outputs data. From the tracker discussion here. With functions to be added instead consumes data, generates data.


 * Because information can not play roles, to indicate that a process has an input of data and an output of data the relations inputs_specified_data and outputs_specified_data have been created.

November 14th 2008
Agenda

Goal is to try and check on current status of the relations branch and have a roadmap for improvement

HouseKeeping

* Welcome to everyone to the resumption of the relations branch calls * We have a new Lead Editor. Frank will be taking over from Liju * Can everyone please send their skype ID to Frank (f.gibson) * Suggestion to move the wiki content of the Relations branch to the new wiki, to be done by Frank if agreed. * The possibility of re-arrange the relations call time, from Friday afternoon. The possible options are: o Monday 4pm GMT o Tuesday 5pm GMT

Issues

Review the proposals circulated by Melanie on the mailing list - Acceptance of these proposals will have a (positive) impact on the numbered issues below.

In addition to the specifics detailed below, those would also get the following annotations:

- curation status "ready for release" - definition editor PERSON: Chris Mungall - definition editor GROUP:OBI: - a preferred term duplicating the rdfs:label

1. inheres_in domain: dependent continuant range: independent continuant

2. addition of inverse relation of inheres_in: bearer_of

def: "A relation between an entity and a dependent continuant; the reciprocal relation of inheres_in" [GOC:cjm] example of usage: red eye bearer_of redness

domain: independent continuant range: dependent continuant

would be added as super property for has_quality, has_function, has_role

3. update has_quality (OBI_0000298) def: "A relation between an entity and a quality. For types: E has_quality Q iff: for any eEt, exists qQt such that q inheres_in e at t. For instances: e has_quality q at t iff q inheres_in e at t and q instance-of Quality" [GOC:cjm] is_a: OBO_REL:bearer_of

4. update has_function (OBI_0000306)

def: "Relation between an independent continuant and a function." [GOC:cjm] is_a: OBO_REL:bearer_of example of usage: heart has_function to-pump-blood

5. update has_role (OBI_0000316) def: "A relation between a continuant C and a role R. The reciprocal relation of role_of." [GOC:cjm] is_a: OBO_REL:bearer_of

6. update is_realization_of (OBI_0000308)

would be renamed "realizes"  with proposed synonyms below

id: OBO_REL:realizes alt_id: OBO_REL:0000034 name: realizes def: "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] synonym: "executes" RELATED [] synonym: "involves_execution_of" RELATED [] synonym: "is_realization_of" EXACT [] synonym: "has_function_part" EXACT [] example of usage: The process of 'histidine catabolism' (GO:0006548) realizes the function 'histidine ammonia lyase activity' (GO:0004397) (note: here 'activity' denotes a function and not a process). We leave open the possibility of defining in future the sub-relations directly_realizes (as bewteen a function and it's functioning) and indirectly_realizes.

7. update is_realized_as (inverse of is_realization_of above) (OBI_0000300)

would be renamed "realized_by" with proposed synonyms below

id: OBO_REL:realized_by alt_id: OBO_REL:realized_as alt_id: OBO_REL:0000035 name: realized_by def: "Relation between a realizable and a process. Reciprocal relation of realizes" [GOC:cjm] synonym: "executed_during" RELATED [] synonym: "has_realization" EXACT []

I/ inverse relation of has_role which has been added to the OBI file as "role is borne by, inverse of OBI_0000316"

1. In order to defined roles and functions each must have a bearer and be borne by an independent continuant and realized during a process. To achieve this we need the following has bearer relations which are children of the has_participant relation. 2. [DS:These role issues are discussed in Michael Dumontiers 'Roles'paper 3. [MC:I think we should consider Robert and Barry's paper Funtion, Role and Disposition in BFO 4. For some reason the relation inheres_in is marked as having an alternative term has_bearer 5. This format could be suitable for qualities. However the range of has_bearer would need to be adjusted as a quality is not a realizable entity.

Option 1:
 * has_bearer <-> is_bearer_of [domain = independent continuant, range = realizable entity]
 * (child) has_role <-> is_role_of [domain = independent continuant, range = role]
 * (child) has_function <-> is_function_of [domain = independent continuant, range = function]

The alternative is

Option 2:
 * has_bearer <-> is_bearer_of [domain = independent continuant, range = realizable entity]
 * (child) has_role_bearer <-> is_bearer_of_role [domain = independent continuant, range = role]
 * (child) has_function_bearer <-> is_bearer_of_function [domain = independent continuant, range = function]

1. In order to defined roles and functions each must have a bearer and be born by an independent continuant and realized during a process. To achieve this we need the following has bearer relations which are children of the has_participant relation. 2. Currently in OBI we have the relation is_realization_of <-> is_realized_as. The inverse of this relation should be is_realized_BY rather than _AS. This is because the realization occurs in or by the realization of a process. Entities such as functions and roles are realized during a process. Realizable entities are participants of a process. Therefore, the relation should be a sub-property of has_participant

has_participant -(child) is_realization_of <-> is_realized_by

III/ has_input and has_output: domain and range are Objective driven process: everything that has_input or has_output is inferred to be an ODP. Proposal from Bjoern to rename those as has_intended_input and has_intended_output. However, there is an alternative proposal from Frank to name the objective_driven_input and objective_driven_output. However, Frank does not believe this should be in as a relation at all, instead we should have the relation is output, which can refer to any output. Then an instance of a output can be described as being objective_driven or not. Currently all this information is being hidden from the reasoner and processes are being classified as objective_driven if they have an output and no objective - this surely is a mistake.

processX has_output some (material and (has_role some objective_driven))