This issue is moved from IEEE Gitlab CCO at https://opensource.ieee.org/cco/CommonCoreOntologies/-/issues/156
It was posted by [Ted Thibodeau Jr]
http://www.ontologyrepository.com/CommonCoreOntologies/biologicalsex arose in w3c/vc-data-model#963
It was part of the description of a human, as --
"biologicalsex": "F"
As I said in the GitHub issue where this arose --
Boy-oh-boy (no pun intended), is biologicalsex) fraught with peril. How are F and whatever alternative values exist defined? I'm pretty sure [you] planned to handle entities assigned "Female" ([typically] XX) or "Male" ([typically] XY) at birth (though I bet [you] don't actually require that the chromosomes be checked), but what about XXY, X, and other variants yet to be discovered? A world of dangers awaits!
@james.schoening pointed me here to continue the discussion....
The current definition of BiologicalSex is
A Quality inhering in a bearer by virtue of the bearer's ability to undergo sexual reproduction in order to differentiate the individuals or types involved.
That "definition" is difficult to decipher, at best, and makes little sense given the value in the example ("F") which seems most likely to stand for "Female" by which you might intend to mean a being which contributes an egg to the reproductive process, as opposed to an "M" which might contribute sperm or similar. But what of a being which may contribute both in a single reproductive action? Or which may contribute one at one time and the other at another time? Or....
Ontologies should always include rdfs:domain and rdfs:range, or at least schema:domainIncludesand schema:rangeIncludes. This is especially true in cases like this, where it appears there is meant to be a limited number of values, but as I said above, there ought to be at least 4 potential values (to handle humans with XX, XY, XXY, and X chromosomes), probably more (to handle other biological entities which may have different chromosomal attributes, and/or be hermaphroditic and/or capable of changing sex characteristics, among other considerations).
Response by
Ronald Rudnicki
@rudnicki
· 1 year ago
Owner
Our definitions of Female Sex and Male Sex were taken from the Phenotypic Trait Ontology. The links in the ontology file are broken, but these should work. https://bioportal.bioontology.org/ontologies/PATO/?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FPATO_0000383 and https://bioportal.bioontology.org/ontologies/PATO/?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FPATO_0000384 .
These classes are subtypes of the Phenotypic Sex. If you look under the Genotypic Sex class I think you'll find consideration of the cases you mention.
I propose that we expand the class of Biological Sex to mirror that of PATO. Please consider that proposal and reply with your judgement.
Response by:
Ted Thibodeau Jr
@macted
· 1 year ago
Author
I'm not sure what you mean by "expand the class of Biological Sex to mirror that of PATO". If you mean, bring the entirety of PATO's Biological Sex, including all contained therein, that would certainly be an improvement. I'll say a bit more, in case that's not what you meant, and also because I think that should be just one of at least a few (probably many) steps, related to all CCO inheritance from external ontologies.
Cherry-picking terms and/or definitions from someone else's ontology is not generally recommended, especially when (as it appears) their ontology is not fully understood. The maintainers of the Phenotypic Quality Ontology (PATO) clearly took great care in its design (even if it remains imperfect), and for reasons that entirely escape me, those who have been working on the Common Core Ontologies apparently decided that only a tiny fraction of PATO was worth inheriting.
At a minimum, I would adjust the definition/description in the CCO to include the details you provided here — especially that your BiologicalSex could probably have been better named PhenotypicSex (and I would recommend that change if your ontology is still sufficiently fluid).
I would also encourage CCO to work with PATO to improve their definitions/descriptions. In immediate focus, "The bearer's ability to undergo sexual reproduction in order to differentiate the individuals or types involved" remains cryptic (indecipherable, really) to me, as "sexual reproduction" does not "differentiate the individuals or types involved"; "sexual reproduction" produces offspring of the individuals participating in that "sexual reproduction".
This issue is moved from IEEE Gitlab CCO at https://opensource.ieee.org/cco/CommonCoreOntologies/-/issues/156
It was posted by [Ted Thibodeau Jr]
http://www.ontologyrepository.com/CommonCoreOntologies/biologicalsex arose in w3c/vc-data-model#963
It was part of the description of a human, as --
"biologicalsex": "F"
As I said in the GitHub issue where this arose --
Boy-oh-boy (no pun intended), is biologicalsex) fraught with peril. How are F and whatever alternative values exist defined? I'm pretty sure [you] planned to handle entities assigned "Female" ([typically] XX) or "Male" ([typically] XY) at birth (though I bet [you] don't actually require that the chromosomes be checked), but what about XXY, X, and other variants yet to be discovered? A world of dangers awaits!
@james.schoening pointed me here to continue the discussion....
The current definition of BiologicalSex is
A Quality inhering in a bearer by virtue of the bearer's ability to undergo sexual reproduction in order to differentiate the individuals or types involved.
That "definition" is difficult to decipher, at best, and makes little sense given the value in the example ("F") which seems most likely to stand for "Female" by which you might intend to mean a being which contributes an egg to the reproductive process, as opposed to an "M" which might contribute sperm or similar. But what of a being which may contribute both in a single reproductive action? Or which may contribute one at one time and the other at another time? Or....
Ontologies should always include rdfs:domain and rdfs:range, or at least schema:domainIncludesand schema:rangeIncludes. This is especially true in cases like this, where it appears there is meant to be a limited number of values, but as I said above, there ought to be at least 4 potential values (to handle humans with XX, XY, XXY, and X chromosomes), probably more (to handle other biological entities which may have different chromosomal attributes, and/or be hermaphroditic and/or capable of changing sex characteristics, among other considerations).
Response by
Ronald Rudnicki
@rudnicki
· 1 year ago
Owner
Our definitions of Female Sex and Male Sex were taken from the Phenotypic Trait Ontology. The links in the ontology file are broken, but these should work. https://bioportal.bioontology.org/ontologies/PATO/?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FPATO_0000383 and https://bioportal.bioontology.org/ontologies/PATO/?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FPATO_0000384 .
These classes are subtypes of the Phenotypic Sex. If you look under the Genotypic Sex class I think you'll find consideration of the cases you mention.
I propose that we expand the class of Biological Sex to mirror that of PATO. Please consider that proposal and reply with your judgement.
Response by:
Ted Thibodeau Jr
@macted
· 1 year ago
Author
I'm not sure what you mean by "expand the class of Biological Sex to mirror that of PATO". If you mean, bring the entirety of PATO's Biological Sex, including all contained therein, that would certainly be an improvement. I'll say a bit more, in case that's not what you meant, and also because I think that should be just one of at least a few (probably many) steps, related to all CCO inheritance from external ontologies.
Cherry-picking terms and/or definitions from someone else's ontology is not generally recommended, especially when (as it appears) their ontology is not fully understood. The maintainers of the Phenotypic Quality Ontology (PATO) clearly took great care in its design (even if it remains imperfect), and for reasons that entirely escape me, those who have been working on the Common Core Ontologies apparently decided that only a tiny fraction of PATO was worth inheriting.
At a minimum, I would adjust the definition/description in the CCO to include the details you provided here — especially that your BiologicalSex could probably have been better named PhenotypicSex (and I would recommend that change if your ontology is still sufficiently fluid).
I would also encourage CCO to work with PATO to improve their definitions/descriptions. In immediate focus, "The bearer's ability to undergo sexual reproduction in order to differentiate the individuals or types involved" remains cryptic (indecipherable, really) to me, as "sexual reproduction" does not "differentiate the individuals or types involved"; "sexual reproduction" produces offspring of the individuals participating in that "sexual reproduction".