obi
obi copied to clipboard
obi-ontology
Add three missing textual definitions
OBI is missing textual definitions for these three terms:
- OBI:0000794 transcription factor binding site
- OBI:0001950 has performer
- OBI:0200001 geometric mean calculation
This is currently the biggest problem with our OBO Dashboard score. @hectorguzor can you please take a first pass at these, or contact the term editor for suggestions?
geometric mean calculation has the following comment by @mbrochhausen
A geometric mean calculation is a descriptive statistics calculation in which the mean is calculated by taking the nth root of the product of all of the observations in a data (n being the number of all observations).
Based on the comment and the following wiki page Geometric_Mean I propose we just use the comment as the definition.
has performer has the following annotation under 'definition editor' [this should probably be changed to a rdfs:comment].
definition editor: performer relation covers the need to report on who performed a planned process. it has to cover processes done by People or Devices (such as a robot controlled by software WF management system).
based on this comment I propose the following definition:
a relation between a planned process and a continuant where the continuant can be a person, organization or device (such as a robot controlled by software workflow management system) that performs the planned process.
Currently ‘has performer’ has range ('Homo sapiens' or organization or device) and ('has role' some 'investigation agent role'). Should this be revised to allow for AIs that can perform various data transformation tasks? If so, should the definition reflect this?
transcription factor binding site is currently a child of sequence feature annotation. ‘sequence feature annotation’ is an ICE defined as
Information about a sequence region.
‘transcription factor binding site' has an annotation that specifies the definition sources as TF_binding_site, defined as
A DNA site where a transcription factor binds
One thing to note is that in SO TF_binding_site (SO:0000235) is under ‘region’ and in OBI transcription factor binding site is under ‘information content entity’. ‘region’ and “information content entity’ are siblings in OBI.
If we use the SO definition, should we move ‘transcription factor binding site’ under 'region'? If we keep ‘transcription factor binding site’ under ICE should the definition be something like:
Information about a DNA site where a transcription factor binds.
For has performer, its good to nail down the semantics of the range a bit, to include "'Homo sapiens' or organization or device", but to use 'has performer' with process modelling in general that often are not about biomedical investigations, I'd like to see the 'has role' some 'investigation agent role' be dropped. It could be added to connect instances by users where necessary.
Discussed in OBI 2022-07-11:
-
define "geometric means calculation" as "A data transformation in which the mean is calculated by taking the nth root of the product of all of the observations in a data (n being the number of all observations)".
-
define "has performs" as "a relation between a planned process and a continuant where the continuant can be a person, organization or device (such as a robot controlled by software workflow management system) that performs the planned process."
-
change what is in under 'definition editor' to 'rdfs:comment'. Drop 'has role' some 'investigation agent role' from range. Should we also drop "'Homo sapiens' or organization or device" to allow for possibility of AI's as task performers? If we do, what do we put for range?
-
there was no discussion on definition for transcription factor binding site. To be discussed in a future OBI call.
Discussed in OBI 2022-08-22: Jie and James think we should keep 'investigation agent role' as is because of the scope of OBI. Perhaps we can change the label to 'has investigation performer' to reflect the textual and logical definition, and request a more general 'has performer' relation in RO?
We also discussed the issue about leaving "'Homo sapiens' or organization or device" in range. James and Hector agreed we should keep it as is for now.
we discussed the definition of 'transcription factor binding site'. We agreed that this term should be deprecated and we should import TF_binding_site (SO:0000235) and use it in its place. 'transcription factor binding site' is currently under ICE, TF_binding_site will be under region.
If 'investigation agent role' is referenced in a renamed 'has investigation performer' relation then we should restrict domain for the relation from 'planned process' to the 'investigation' process instead? I'm looking up 'investigation agent role' and can see it only makes reference to investigation, and never to 'planned process'. I didn't realize it had so many subclassed roles.
At any rate I'm fine with a new generic 'has performer' role if existing one is relabeled/redefined.
First thoughts:
geometric mean calculation is easy: what is the comment should be the definition 'has performer' should probably live in COB/RO, but for now we should add a definition like: "A performer is a person or device that is carrying out the action specifications in a planned process" : transcription factor binding site was supposed to be a placeholder for an SO term apparently; maybe we should obsolete and replace with an important SO term instead?
On Mon, Jun 6, 2022 at 12:47 PM James A. Overton @.***> wrote:
OBI is missing textual definitions for these three terms:
- OBI:0000794 transcription factor binding site http://purl.obolibrary.org/obo/OBI_0000794
- OBI:0001950 has performer http://purl.obolibrary.org/obo/OBI_0001950
- OBI:0200001 geometric mean calculation http://purl.obolibrary.org/obo/OBI_0200001
This is currently the biggest problem with our OBO Dashboard score. @hectorguzor https://github.com/hectorguzor can you please take a first pass at these, or contact the term editor for suggestions?
— Reply to this email directly, view it on GitHub https://github.com/obi-ontology/obi/issues/1553, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2IWZ4VV64FAJX6TLGB3VNZIW3ANCNFSM5YAPW44Q . You are receiving this because you are subscribed to this thread.Message ID: @.***>
-- Bjoern Peters Professor La Jolla Institute for Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters
A PR that addresses the missing definitions has been merged. So I will close this issue. @ddooley We should address the concerns with has performer you brought up here. But we should address these in a new issue.