Re: [Obo-taxonomy] [Obo-discuss] Proposal for a rank ontology (vocabulary)
Barry Smith
It would be nice to see a definition of 'taxonomic rank', since this
plays so overwhelming a role in the ontology.
[Typedef]
id: TAXRANK:has_rank
name: has_rank
alt_id: TAXRANK:0000033
comment: A metadata relation more or less equivalent to the similar
relation in NCBI taxonomy.
is_metadata_tag: true
I would like to see a definition of this relation following the rules
set out in
http://genomebiology.com/2005/6/5/R46
You have many assertions of the form
tribe is_a taxonomic_rank
The recommended definition of 'is_a' tells us that
if A is_a B, then every instance of A is an instance of B.
Thus you are asserting that for every x,
if x instance_of tribe, then x instance_of taxonomic rank.
This does not seem right. Does this mean that you have another
definition of is_a? Do you mean by 'is_a' perhaps 'instance_of'? If
so, then you should say so. Then, however, we have not an ontology of
ranks but rather, like GAZ, a controlled vocabulary built upon
ontological principles. This is because ontologies represent only
types, and you are representing just one type, together with
providing a list of instances.
BS
At 05:16 PM 5/28/2009, Peter E. Midford wrote:
>Hello,
> After over a year of discussing alternative approaches to
>representing taxonomic ranks, I am submitting a proposal for an
>ontology of taxonomic ranks that closely mirrors the way these are
>implemented in the NCBI and Teleost taxonomy ontologies. This
>ontology is a break-out of the rank terms that appear in these
>taxonomy ontologies, allowing both ontologies, and others, to have a
>single root for their terms.
>
>As many of you have heard my proposals for representing taxa as
>individuals, this design may come as a surprise. There are both
>pragmatic and philosophical reasons for this choice. Pragmatically,
>the current OBO tools (e.g., OBO Edit) are not setup to treat
>instances as first class entities. Although I have discussed OBO-Edit
>add-ons for converting taxonomies between instances and terms, storing
>taxonomies as hierarchies of individuals would complicate use or reuse
>by other projects. As several other projects have used the existing
>TTO and NCBI taxonomy ontologies as models, changing TTO would
>potentially split the community. Philosophically, I have come to see
>that taxonomy ontologies are better considered as collections of
>entities called 'taxon concepts' - information entities that appear in
>publications - rather than clades. If taxonomy ontologies are about
>published entities rather than clades, than most of the arguments for
>the terms being metaphysical individuals rather than classes (e.g.,
>those of Ghiselin) no longer apply.
>
>The attached obo file contains a set of taxonomic rank terms that
>includes the ranks from the NCBI taxonomy as well as a few additional
>ones I picked up from reviewing the Zoological and Botanical codes.
>I'll admit to a bit of zoological chauvinism in assigning phylum as a
>primary term and division as a synonym, but as the plant ontology
>people seem to have their own system for coding ranks in their
>ontology, this preference for zoological terms reflects the majority
>of the use community for this ontology. I have also removed any
>ordering relation as having one seems to cause more trouble than it is
>worth.
>
>My apologies for having held this so long. I hope we can quickly
>resolve this issue, either with this set or an alternative.
>
>Thanks,
>
>Peter
>
>
>
>
>
>
>
>
>
>--
>Peter Midford
>Phenoscape Taxonomy Curator
>
>
>
>
>
>
>------------------------------------------------------------------------------
>Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
>is a gathering of tech-side developers & brand creativity professionals. Meet
>the minds behind Google Creative Lab, Visual Complexity, Processing, &
>iPhoneDevCamp as they present alongside digital heavyweights like Barbarian
>Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com
>_______________________________________________
>Obo-discuss mailing list
>Obo-d...@lists.sourceforge.net
>https://lists.sourceforge.net/lists/listinfo/obo-discuss
------------------------------------------------------------------------------
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
is a gathering of tech-side developers & brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing, &
iPhoneDevCamp as they present alongside digital heavyweights like Barbarian
Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com
_______________________________________________
Obo-taxonomy mailing list
Obo-ta...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/obo-taxonomy
Peter Midford
Some comments
It would be nice to see a definition of 'taxonomic rank', since this
plays so overwhelming a role in the ontology.
[Typedef]
id: TAXRANK:has_rank
name: has_rank
alt_id: TAXRANK:0000033
comment: A metadata relation more or less equivalent to the similar
relation in NCBI taxonomy.
is_metadata_tag: true
I would like to see a definition of this relation following the rules
set out in
http://genomebiology.com/2005/6/5/R46
You have many assertions of the form
tribe is_a taxonomic_rank
The recommended definition of 'is_a' tells us that
if A is_a B, then every instance of A is an instance of B.
Thus you are asserting that for every x,
if x instance_of tribe, then x instance_of taxonomic rank.
This does not seem right. Does this mean that you have another
definition of is_a? Do you mean by 'is_a' perhaps 'instance_of'? If
so, then you should say so. Then, however, we have not an ontology of
ranks but rather, like GAZ, a controlled vocabulary built upon
ontological principles. This is because ontologies represent only
types, and you are representing just one type, together with
providing a list of instances.
BS
_______________________________________________
Phenoscape mailing list
Pheno...@nescent.org
https://lists.nescent.org/mailman/listinfo/phenoscape
Barry Smith
>Barry,
>>Some comments
>>
>>It would be nice to see a definition of 'taxonomic rank', since this
>>plays so overwhelming a role in the ontology.
>
>Agree,
>
>Here's the current definition in the NCBI taxonomy ontology
>"A level of depth within a species taxonomic tree"
>
>Definition in ICZN glossary (of rank)
>"the level, for nomenclatural purposes, of a taxon in a taxonomic
>hierarchy(e.g. all families are for nomenclatural purposes at the
>same rank, which lies between superfamily and subfamily)."
>
>I would suggest combining these into: A level of depth of a taxon in
>a taxonomic hierarchy. This avoids the specification of a 'species
>taxonomic tree,' since a taxonomy need neither extend down to the
>species level, nor stop there (NCBI has terms below the species level already).
I agree with this; however, do all of your examples really fit this
definition? do not some of them span multiple levels, for example?
>>[Typedef]
>>id: TAXRANK:has_rank
>>name: has_rank
>>alt_id: TAXRANK:0000033
>>comment: A metadata relation more or less equivalent to the similar
>>relation in NCBI taxonomy.
>>is_metadata_tag: true
>>
>>I would like to see a definition of this relation following the rules
>>set out in
>><http://genomebiology.com/2005/6/5/R46>http://genomebiology.com/2005/6/5/R46
>
>NCBI currently has
>
> a metadata relation between a class and its taxonomic rank (e.g.,
> species, family)
>
>This doesn't seem sufficient to me. If we treating taxonomic
>identifiers as classes, then to avoid metaclasses, we should at
>least specify a root superclass for all these identifiers. This
>root could be defined in this ontology, then any taxonomic ontology
>using this relation would specify the root specified here as the
>parent of whatever they use for a root (e..g, Chordata).
>
>So perhaps "a metadata relation between a subclass of TAXRANK:root
>and a subclass of TAXRANK:taxonomic_rank, that specifies the depth
>of the former in the taxonomic tree"
relations normally do not specify; in general the subject of the verb
'to specify' should be a datum, I think -- so this definition needs
rephrasing (but do we really need such an odd relation at all?)
>>You have many assertions of the form
>>
>>tribe is_a taxonomic_rank
>>
>>The recommended definition of 'is_a' tells us that
>>
>>if A is_a B, then every instance of A is an instance of B.
>>
>>Thus you are asserting that for every x,
>>
>>if x instance_of tribe, then x instance_of taxonomic rank.
>>
>>This does not seem right. Does this mean that you have another
>>definition of is_a? Do you mean by 'is_a' perhaps 'instance_of'? If
>>so, then you should say so. Then, however, we have not an ontology of
>>ranks but rather, like GAZ, a controlled vocabulary built upon
>>ontological principles. This is because ontologies represent only
>>types, and you are representing just one type, together with
>>providing a list of instances.
>>BS
>
>Except that none of these terms (expect rank itself) have is_a
>children - the only children they have are via the has_rank
>relation, which as a metadata relation supports no inference at
>all. Doing it this way seems to work, but the design decision
>(making has_rank a metadata relation) was originally made by Chris
>(I believe).
I think we need to involve Chris in this discussion
BS
>>><http://p.sf.net/sfu/creativitycat-com>http://p.sf.net/sfu/creativitycat-com
>>>_______________________________________________
>>>Obo-discuss mailing list
>>><mailto:Obo-d...@lists.sourceforge.net>Obo-d...@lists.sourceforge.net
>>>https://lists.sourceforge.net/lists/listinfo/obo-discuss
>>
>>
>>_______________________________________________
>>Phenoscape mailing list
>><mailto:Pheno...@nescent.org>Pheno...@nescent.org
>>https://lists.nescent.org/mailman/listinfo/phenoscape
>
>Peter E. Midford
>Phenoscape Taxonomy Curator
><mailto:Peter....@gmail.com>Peter....@gmail.com
>
>
------------------------------------------------------------------------------
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
is a gathering of tech-side developers & brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing, &
iPhoneDevCamp as they present alongside digital heavyweights like Barbarian
Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com
_______________________________________________
Pankaj Jaiswal (OSU)
How about using the rank names as relationship types. If so then we do
not have to create a separate ranking system with 'rank' classes that
have taxonomic names as children. The relationship_types can have
rankings of their own.
If I go by later we may be proposing something like
[i] = is_a
Oryzeae
--[i] oryza
----[i] oryza sativa
taxonomic rank
--[i] tribe
----[i] Oryzeae
--[i] genus
----[i] oryza
--[i] species
----[i] oryza sativa
However if we go by former by using ranks as relationship types, then it
is more clear inference as follows and no need to extend the data
structure to populate the rank classes.
Oryzeae
--[is_genus_of] oryza
----[is_species_of] Oryza sativa
Also what about the interspecific terms that we have many in plants (not
all listed on NCBI)
genetic cross of two taxons, e.g.
name: Triticum aestivum x Secale
synonym: Secale cereale x Triticum
aestivumcerealehttp://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?mode=Info&lvl=3&keep=1&srchmode=1&unlock&id=49318
Terms like these can have two parents (the two parent species) may be
with a different relationship_type (e.g. is_intersection_of
/interspecific_cross_of) other than species used by NCBI.
------------------------------------------------------------------------------
OpenSolaris 2009.06 is a cutting edge operating system for enterprises
looking to deploy the next generation of Solaris that includes the latest
innovations from Sun and the OpenSource community. Download a copy and
enjoy capabilities such as Networking, Storage and Virtualization.
Go to: http://p.sf.net/sfu/opensolaris-get
Peter Midford
I'm happy to see a Botanical perspective on this, since
it would be best to come up with a solution that works for everyone.
My two questions about treating ranks as relationship types are: i)
how to define them to capture the intended meaning - making them
simple subrelations of is_a, with no differentia won't go over with
Barry I expect; ii) how to reason with them. I don't know if you do
any reasoning using your taxonomy, but in Phenoscape we decided that
trying to avoid using taxonomic inheritance is impractical. I doubt
we are the only project that has been led back to taxonomic
inheritance for practical reasons. So, if you want to inherit
properties from Oryzeae down to Oryza sativa, than rather than saying
that is_a is transitive, you would need to define inheritance for
compositions (e.g., is_species_of compose is_genus_of) to infer that
O. sativa inherits from Oryzeae.
I'd be happy to see solutions to these, as it sounds like an
interesting approach.
Cheers,
Peter
On Jun 1, 2009, at 15:13, Pankaj Jaiswal (OSU) wrote:
Peter E. Midford
Phenoscape Taxonomy Curator
Peter....@gmail.com
------------------------------------------------------------------------------
OpenSolaris 2009.06 is a cutting edge operating system for enterprises
looking to deploy the next generation of Solaris that includes the latest
innovations from Sun and the OpenSource community. Download a copy and
enjoy capabilities such as Networking, Storage and Virtualization.
Go to: http://p.sf.net/sfu/opensolaris-get
Pankaj Jaiswal (OSU)
I agree with your thought on not to avoid the taxonomic inheritance. I
am sure Todd Vision is following this as he is involved in some of the
similar reasoning projects on the iPlant. There is more involvement of
iPlant in other projects e.g on phenology, where we need these
reasonings to be built and I am participating in them. In the meantime
as I have seen, often more granular terms in taxonomy edges belonging to
rank species and below are most detailed and are somewhat fixed. Whereas
their parentage nodes may change as we learn more about the biological
systems and phylogenomics using molecular genetic and other approaches.
Your transitive approach seems great as a wholesome approach, however,
often we can say with confidence in an intransitive fashion what do O.
sativa and its siblings contribute to Oryzeae to make it unique as
compared to what is unique about Oryzae that is always inherited
transitively in all the siblings for example Oryza, Zizania, and
Leersia. We know everything is not inherited transitively in this case.
Pankaj
>> /interspecific_cross_of) other than species used by NCBI.
>
> Peter E. Midford
> Phenoscape Taxonomy Curator
> Peter....@gmail.com
>
>
>
>
--
Pankaj Jaiswal
Assistant Professor
Dept. of Botany and Plant Pathology
3082 Cordley Hall
Oregon State University
Corvallis, OR, 97331-2902, USA
Ph.: +1-541-737-8471
Fax: +1-541-737-3573
Web: www.gramene.org
www.plantontology.org
Erick Antezana
could you elaborate a bit more on the "reasoning" scenarios you mentioned? (competency questions, use cases,...)
thanks,
Erick
Chris Mungall
I'm listening, but don't have anything much to add to Peter's summary
- ranks are not types in the sense used by BFO, RO etc. They are just
convenient terms that are used to indicate depth in a taxonomy
- has_rank is simply a way of associating a class in a taxonomy with
its rank
- the scheme is deliberately ontologically and logically weak, like,
for example, GO slims.
This appears to be work for the evolutionary biologists for now,
according to Peter. At some point they may wish to move to a system
that makes stronger ontological claims regarding species, clades etc
and the relations that hold between them. I think this could live
within the existing framework, or side-by-side without contradiction.
> Obo-discuss mailing list
> Obo-d...@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/obo-discuss
>
------------------------------------------------------------------------------
OpenSolaris 2009.06 is a cutting edge operating system for enterprises
looking to deploy the next generation of Solaris that includes the latest
innovations from Sun and the OpenSource community. Download a copy and
enjoy capabilities such as Networking, Storage and Virtualization.
Go to: http://p.sf.net/sfu/opensolaris-get
Pankaj Jaiswal (OSU)
-What are the morphological and developmental landmarks that e.g.
various Oryza species ( sativa, galberrima, etc) and its siblings share
and/or differ on.
- I presume the shared characters can them be the landmarks for higher
order genus or any parents in the tree. If we go by the current
practices of annotations the parent terms will obviously acquire both
the common and uncommon features. However, the uncommon features will
not be applicable intransitively since not all the children will have
all those feature.
-If a (set of) gene(s) displays phenotype/function/expression for
example in O. sativa species, what's the likelihood of the orthologous
gene displaying the same phenotype/function/expression in sibling
species or a related species in a phylogenetic tree but coming from a
different genus of the common tribe/phyla.
-How does these set of orthologous genes evolved in different species
(or even accessions of the species) and contribute to genetic diversity,
response to environment/envO (biotic and abiotic) and georeference site
(GAZ), loss/gain of function and paralogous pairs that may also involve
ploidy or genome duplication.
- some may also want to study the order of genes in syntenic region
between the two species
- extend it to study of regulatory cis and trans elements of the
orthologous genes.
Pankaj
Erick Antezana wrote:
> Hi Pankaj,
>
> could you elaborate a bit more on the "reasoning" scenarios you
> mentioned? (competency questions, use cases,...)
>
> thanks,
> Erick
>
> 2009/6/1 Pankaj Jaiswal (OSU) <jais...@science.oregonstate.edu
> <mailto:jais...@science.oregonstate.edu>>
> > Peter....@gmail.com <mailto:Peter....@gmail.com>
> >
> >
> >
> >
>
> --
> Pankaj Jaiswal
> Assistant Professor
> Dept. of Botany and Plant Pathology
> 3082 Cordley Hall
> Oregon State University
> Corvallis, OR, 97331-2902, USA
>
> Ph.: +1-541-737-8471
> Fax: +1-541-737-3573
> Web: www.gramene.org <http://www.gramene.org>
> www.plantontology.org <http://www.plantontology.org>
>
>
> ------------------------------------------------------------------------------
> OpenSolaris 2009.06 is a cutting edge operating system for enterprises
> looking to deploy the next generation of Solaris that includes the
> latest
> innovations from Sun and the OpenSource community. Download a copy and
> enjoy capabilities such as Networking, Storage and Virtualization.
> Go to: http://p.sf.net/sfu/opensolaris-get
> _______________________________________________
> Obo-discuss mailing list
> Obo-d...@lists.sourceforge.net
> <mailto:Obo-d...@lists.sourceforge.net>
> https://lists.sourceforge.net/lists/listinfo/obo-discuss
>
>
>
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------------
> OpenSolaris 2009.06 is a cutting edge operating system for enterprises
> looking to deploy the next generation of Solaris that includes the latest
> innovations from Sun and the OpenSource community. Download a copy and
> enjoy capabilities such as Networking, Storage and Virtualization.
> Go to: http://p.sf.net/sfu/opensolaris-get
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Obo-discuss mailing list
> Obo-d...@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/obo-discuss
--
Pankaj Jaiswal
Assistant Professor
Dept. of Botany and Plant Pathology
3082 Cordley Hall
Oregon State University
Corvallis, OR, 97331-2902, USA
------------------------------------------------------------------------------
OpenSolaris 2009.06 is a cutting edge operating system for enterprises
looking to deploy the next generation of Solaris that includes the latest
innovations from Sun and the OpenSource community. Download a copy and
enjoy capabilities such as Networking, Storage and Virtualization.
Go to: http://p.sf.net/sfu/opensolaris-get
_______________________________________________
Roger Hyam
I have no idea which of these mailing lists I am on so this will
probably bounce several times....Perhaps some one could repost this to
the correct lists.
I/We did quite a lot of thinking about this for the TDWG ontology/
vocabulary stuff over the past few years but I am not familiar with
the other OBO based onotologies mentioned.
These TDWG ontologies are currently under review again. We have an OWL
ontology of ranks
http://rs.tdwg.org/ontology/voc/TaxonRank
where the ranks are treated as individuals.
We used to have the ability for both taxa and names to have ranks
associated with them but then deprecated the notion of rank from
TaxonConcept leaving it as purely a nomenclatural notion.
http://rs.tdwg.org/ontology/voc/TaxonConcept
http://rs.tdwg.org/ontology/voc/TaxonName
My personal thoughts at the moment are that ranks bear little or no
semantic meaning or value. Knowing something is a species for example
one can conclude very little else about it that one could conclude if
it was any other rank (higher and lower taxa for example). Just about
the only use is checking that names are well formed and rendered but
the logic involved in doing this is complex and a one off job so it is
probably better for a human to do it and not to bother modeling all
the complexities of the ICBN and ICZN in OWL or some other ontology
language.
I currently believe, for our purposes, that taxa are classes and taxon
names are the equivalent of OWL class annotations. Ranks are therefore
like property values of annotations which gives an indication of the
biological significance!
Hope this is a helpful contribution.
Roger
Hilmar Lapp
On Jun 5, 2009, at 5:13 AM, Chris Mungall wrote:
> - ranks are not types in the sense used by BFO, RO etc. They are just
> convenient terms that are used to indicate depth in a taxonomy
> - has_rank is simply a way of associating a class in a taxonomy with
> its rank
> - the scheme is deliberately ontologically and logically weak, like,
> for example, GO slims.
>
> This appears to be work for the evolutionary biologists for now,
> according to Peter. At some point they may wish to move to a system
> that makes stronger ontological claims regarding species, clades etc
> and the relations that hold between them.
Yes. It's worth keeping in mind that ranks in taxonomy are primarily a
classification aid and as such are very helpful.
As for reasoning with ranks, defining ranks phylogenetically, or by
genetic or character distance or cohesion is a matter of perennial and
sometimes heated debate, and I think misses the difference between
taxonomy and phylogeny. Despite all this, the ability to classify life
on earth in a (mostly) coherent and hierarchical system with named
node depths has proven extremely useful over the last 200 years.
You may be aware that there are also efforts at rank-free taxonomy
(PhyloCode, for example), but the vast majority of (contemporary)
taxonomists insist that that will never catch on ...
-hilmar
--
===========================================================
: Hilmar Lapp -:- Durham, NC -:- hlapp at duke dot edu :
===========================================================
Hilmar Lapp
ontology to help with or be applied towards those unless you include
phylogenetic definitions (which I believe would be highly contentious
as these would necessarily be hypotheses, not what we know about
reality).
-hilmar
> _______________________________________________
> Phenoscape mailing list
> Pheno...@nescent.org
> https://lists.nescent.org/mailman/listinfo/phenoscape
--
===========================================================
: Hilmar Lapp -:- Durham, NC -:- hlapp at duke dot edu :
===========================================================
------------------------------------------------------------------------------
Hilmar Lapp
On Jun 5, 2009, at 2:55 PM, Roger Hyam wrote:
> These TDWG ontologies are currently under review again. We have an OWL
> ontology of ranks
>
> http://rs.tdwg.org/ontology/voc/TaxonRank
>
> where the ranks are treated as individuals.
This is great, and is in OWL format if I understand correctly (and
seems more comprehensive than the list we/Peter came up with).
I think (though I'm not sure) it is rather difficult (if not currently
unsupported) to import terms from non-OBO ontologies into OBO
ontologies (which is what we would want to do with the Taxon Rank
ontology). (Can someone of one of the OBO lists with insight into the
OBO format and OBO-Edit inner workings confirm or reject this?)
Would you (assuming you are the maintainer?) be adverse to submitting
this ontology to the OBO Foundry? (Inclusion in the OBO Foundry [1]
among other things means agreeing to a number of design, editorial,
and maintenance principles [2]. See also the Smith et al, Nat Biotech
article [3]. The biggest issue would probably be globally unique
opaque identifiers for each term, which OBO Foundry requires.)
Peter - do you think trying to reconcile with the TDWG TaxonRank
ontology and submitting the result to OBO Foundry would be more
trouble than it'd be worth?
-hilmar
[1] http://obofoundry.org/
[2] http://obofoundry.org/wiki/index.php/OBO_Foundry_Principles
[3] http://dx.doi.org/10.1038/nbt1346
--
===========================================================
: Hilmar Lapp -:- Durham, NC -:- hlapp at duke dot edu :
===========================================================
------------------------------------------------------------------------------
Pankaj Jaiswal
That's why in the first place I suggested of using the ranks as relationships, instead of
taxon-X | has_rank | rank-Y (e.g. O. sativa | has_rank | species). With my proposal one can build the phylogenomic/a taxonomy tree in its simplest of the form. See more below. Alternatively if we go by current suggestion of taxon-X has_rank rank-Y strategy, one will end up having ranks as classes, but children of those classes will be nothing but a long flat list of terms, e.g. imagine more than million species terms as children of rank_species, which may not resolve any of the following questions I had.
Regarding the multiple phylogenetic trees, there is a diverse array of trees and not many consensus ones. Most of the studies are done using certain molecular and morphological markers which may or may not represent the structure and organization of the whole genomes. If the whole genomes are sequenced** and aligned, we may see a different phylogeny compared to certain loci which may have their won localized phylogeny depending on how actively evolving is that region of the genome in a given species. So if we traverse the whole genome based phylogeny we may see it breaking it a certain points compared to the whole genome. I am sure as we sequence more genomes this will be the method of choice.
Conventional taxonomy, on the other hand followed by users and text books use Linnaean taxonomy that is taught from K-12 and even in colleges and is widely used. The question comes to a point where we see diverse sources of taxonomy and many using their own version as we have seen people mentioning it in this thread Not to forget 'no_rank in NCBI's taxonomy version. Besides ontology use the consensus can be built among various versions, using the similar strategy described by Yap et al (Genetics, Vol. 165, 2235-2247, December 2003 http://www.genetics.org/cgi/content/abstract/165/4/2235) for genetic maps, where we see a common problem of diverse position/order (top to bottom) on a given genetic map. Analogous to taxonomy which is one single genetic map with taxons ordered from top to bottom.
Let me know if it helps.
Pankaj
** Whole genome sequencing is not necessarily a representation of the whole species, instead a version of its genome based on the strain that was sequenced. A strain may have a gene that is knocked out / truncated or even a null allele, it may still put the phylogenoy out of place, and will not be a true representation in the tree
Judith Blake
judy
Hilmar Lapp
Nagarjuna G.
context fully. While reading the posts, I came across some issues
raised, one a possible ambiguous use of is_a, and another an attempt
not to use metaclasses (metatypes) while representing taxonomic
ranks.
I do not know how one could avoid the use of metatypes while
involving
taxonomic ranks.
One way of clearing the issue could be by avoiding the use of the
expression is_a and use the alternate technical term "member of" in
its place or use is_a only as a prefix to the following two
relations. Let us clearly distinguish the two basic but completely
different relations:
1. is a member of (same as instance of): this relation holds only
between
a token (individual) and a type.
2. is a type of (same as subtype of): this relation holds only
between two
types (classes)
1 is a non-transitive relation, while 2 is transitive.
As ontologists "is_a" can be eliminated completely, for people tend
to
use it (at least in folklore as in "Nagarjuna is an animal"), for
expressing both the relations 1 and 2. If one wants to use, use it
as
a prefix to the exact intended meaning.
All the taxonomic ranks hold between them the relation 2 or its
inverse 'super type of'.
Following the above suggestion: we shouldn't say "Mollusca is a
phylum", instead we say in unambiguous terms "Mullusca is a member of
phylum". Though this relation is between two types we need to
distinguish the levels, where the former (Mollusca) is a type and the
latter (phylum) is metatype, where types are tokens (individuals) of
metatypes. I do not think taxonomy can be done without using
metatypes. Taxonomic ranks belongs to metatypes. "types of tokens
of
metatypes" causes some confusion to begin with, but when we apply it
as a rule, it is rigorous and causes no confusion.
Similarly when we have to say "Snail is a Molluscan", we say instead
"Snail is a type of Mullusca", for Snail is also a type. Thus between
Snail and Mollusca relation 2 holds.
Thus relation 1 can be used in an ontology only between a type and
and
a metatype. It cannot be used between two types or between two
metatypes. Therefore it cannot be used between two taxonomic ranks.
At another _lower_ level, seldom used while defining ontologies,
"Nagarjuna is a chordate" can be expressed as "Nagarjuna is a member
of chordate". If one tends to use is_a every where, we may obtain by
inference "Nagarjuna is_a phylum", which is wrong.
(Wrong inference: Nagarjuna is a chordate, chordata is a
phylum. therefore Nagarjuna is a phylum. In this argument is_a in the
antecedents stands for two different relations, that is why leads to
a
wrong conclusion.)
This possibility can be broken only if we ensure correct uses of 1
and
2, and also the use of metatypes. I don't think the issue can be
resolved without using metatypes.
Hope my post is not orthogonal to the issue being discussed.
--
Nagarjuna G.
gnowledge lab, HBCSE, TIFR. India.
------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables unlimited
royalty-free distribution of the report engine for externally facing
server and web deployment.
http://p.sf.net/sfu/businessobjects
Nagarjuna G.
>
> Following the above suggestion: we shouldn't say "Mollusca is a
> phylum", instead we say in unambiguous terms "Mullusca is a member
> of phylum". Though this relation is between two types we need to
> distinguish the levels, where the former (Mollusca) is a type and
> the latter (phylum) is metatype, where types are tokens
> (individuals) of metatypes. I do not think taxonomy can be done
> without using metatypes. Taxonomic ranks belongs to metatypes.
> "types of tokens of
> metatypes" causes some confusion to begin with, but when we apply
> it as a rule, it is rigorous and causes no confusion.
>
In the above passage please read "types of tokens of metatypes" as
"types ARE tokens of metatypes".
sorry for the typo.