Organism Taxonomic Ontology

classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|

Organism Taxonomic Ontology

Pete DeVries
I am working to create a taxonomic ontology with protege.
I am not sure how to treat these related
concepts.

Kingdom
  Phylum
    Subphylum
      Class
        Subclass
          Infraclass
             Order
               Suborder
                 Infraorder
                    Family
                      Subfamily
                         Tribe
                            Genus
                                specific_epithet

and the hierarchy for a specific organism

Kingdom Animalia
  Phylum Arthropoda
    Subphylum Hexapoda
      Class Hexapoda
        Subclass Insecta
          Infraclass Neoptera
             Order Diptera
               Suborder Nematocera
                 Infraorder Culicomorpha
                    Family Culicidae
                      Subfamily Anophelinae
                         Tribe     "tribe unassigned"
                            Genus Anopheles
                                specific_epithet barberi

For some groups of organisms levels like "Tribe" are commonly used,
while the are not used for other groups. My idea is is to use "tribe_unassigned"
etc. when a given level is not used.

I am thinking that it would be best to have two ontologies the first
being the simple taxonomic hierarchy and the second that incorporates
this ontology while adding the taxon information.

I am also not sure how to show that there are several instances within
a sublevel for instance 5 kingdoms and Animalia is just one, and Arthropoda
is a subclass of Animalia.

I looked around for existing ontologies but I did not find one that
seemed complete. Also in the Kingdom Plantae the level Division is
used rather than Phylum, otherwise it is the same.

Thanks in Advance :-)

- Pete

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Organism Taxonomic Ontology

Steve Wartik
Pete,

It sounds as if you are dealing with the classes-as-individuals issue. Have you read http://www.w3.org/TR/swbp-classes-as-values/?

Steve Wartik

Pete DeVries wrote:
I am working to create a taxonomic ontology with protege.
I am not sure how to treat these related
concepts.

Kingdom
  Phylum
    Subphylum
      Class
        Subclass
          Infraclass
             Order
               Suborder
                 Infraorder
                    Family
                      Subfamily
                         Tribe
                            Genus
                                specific_epithet

and the hierarchy for a specific organism

Kingdom Animalia
  Phylum Arthropoda
    Subphylum Hexapoda
      Class Hexapoda
        Subclass Insecta
          Infraclass Neoptera
             Order Diptera
               Suborder Nematocera
                 Infraorder Culicomorpha
                    Family Culicidae
                      Subfamily Anophelinae
                         Tribe     "tribe unassigned"
                            Genus Anopheles
                                specific_epithet barberi

For some groups of organisms levels like "Tribe" are commonly used,
while the are not used for other groups. My idea is is to use "tribe_unassigned" 
etc. when a given level is not used.

I am thinking that it would be best to have two ontologies the first
being the simple taxonomic hierarchy and the second that incorporates
this ontology while adding the taxon information.

I am also not sure how to show that there are several instances within
a sublevel for instance 5 kingdoms and Animalia is just one, and Arthropoda
is a subclass of Animalia.

I looked around for existing ontologies but I did not find one that
seemed complete. Also in the Kingdom Plantae the level Division is
used rather than Phylum, otherwise it is the same.

Thanks in Advance :-)

- Pete

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html



  

Reply | Threaded
Open this post in threaded view
|

Re: Organism Taxonomic Ontology

Pete DeVries
Steven Wartik <swartik <at> ida.org> writes:

>
>
> Pete,
> It sounds as if you are dealing with the classes-as-individuals issue.
> Have you read http://www.w3.org/TR/swbp-classes-as-values/?
> Steve WartikPete DeVries wrote:


I read through this and it not clear to me what option is best
for my implementation.

I am now wondering if I should have a separate owl for each
species which has the species full taxonomic hierarchy
and then use someway to point to the classes
so that it is clear that a specific group
 .. lets say .. Arthropoda in one owl file
is the same as Arthropoda in another owl file.

Does this make sense?

Thanks in advance :-)

- Pete

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Organism Taxonomic Ontology

samsontu

One approach is to use metaclasses for the hierarchy of Kingdom, Phylum,
Subphylum, etc.

The Animalia, Arthropoda, etc. will be instances of Kingdom, Phylum,
Subphylum respectively and form their own hierarchy.

The issue of optional Tribe may be handled by constraints (restrictions)

SAmson



Pete DeVries wrote:

> Steven Wartik <swartik <at> ida.org> writes:
>
>
>>
>>Pete,
>>It sounds as if you are dealing with the classes-as-individuals issue.
>>Have you read http://www.w3.org/TR/swbp-classes-as-values/?
>>Steve WartikPete DeVries wrote:
>
>
>
> I read through this and it not clear to me what option is best
> for my implementation.
>
> I am now wondering if I should have a separate owl for each
> species which has the species full taxonomic hierarchy
> and then use someway to point to the classes
> so that it is clear that a specific group
>  .. lets say .. Arthropoda in one owl file
> is the same as Arthropoda in another owl file.
>
> Does this make sense?
>
> Thanks in advance :-)
>
> - Pete
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>
>


--
Samson Tu                    email: [hidden email]
Senior Research Scientist    web: www.stanford.edu/~swt/
Stanford Medical Informatics phone: 1-650-725-3391
Stanford University          fax: 1-650-725-7944

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Organism Taxonomic Ontology

Alan Ruttenberg-2
In reply to this post by Pete DeVries
Hi Pete.

One thing you might think about is what inferences/queries do you want
to be able to do nand what constraints do you want automaticallly
checked? Given those, it would be easier to evaluate a particular
proposal.

BTW, how would you represent "mussels" - this came up for me in a
problem I am working on .

-Alan

On 5/10/06, Pete DeVries <[hidden email]> wrote:

> I am working to create a taxonomic ontology with protege.
> I am not sure how to treat these related
> concepts.
>
> Kingdom
>   Phylum
>     Subphylum
>       Class
>         Subclass
>           Infraclass
>              Order
>                Suborder
>                  Infraorder
>                     Family
>                       Subfamily
>                          Tribe
>                             Genus
>                                 specific_epithet
>
> and the hierarchy for a specific organism
>
> Kingdom Animalia
>   Phylum Arthropoda
>     Subphylum Hexapoda
>       Class Hexapoda
>         Subclass Insecta
>           Infraclass Neoptera
>              Order Diptera
>                Suborder Nematocera
>                  Infraorder Culicomorpha
>                     Family Culicidae
>                       Subfamily Anophelinae
>                          Tribe     "tribe unassigned"
>                             Genus Anopheles
>                                 specific_epithet barberi
>
> For some groups of organisms levels like "Tribe" are commonly used,
> while the are not used for other groups. My idea is is to use "tribe_unassigned"
> etc. when a given level is not used.
>
> I am thinking that it would be best to have two ontologies the first
> being the simple taxonomic hierarchy and the second that incorporates
> this ontology while adding the taxon information.
>
> I am also not sure how to show that there are several instances within
> a sublevel for instance 5 kingdoms and Animalia is just one, and Arthropoda
> is a subclass of Animalia.
>
> I looked around for existing ontologies but I did not find one that
> seemed complete. Also in the Kingdom Plantae the level Division is
> used rather than Phylum, otherwise it is the same.
>
> Thanks in Advance :-)
>
> - Pete
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>
>
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Organism Taxonomic Ontology

Pete DeVries
Alan Ruttenberg <alanruttenberg <at> gmail.com> writes:

>
> Hi Pete.
>
> One thing you might think about is what inferences/queries do you want
> to be able to do nand what constraints do you want automaticallly
> checked? Given those, it would be easier to evaluate a particular
> proposal.
>
> BTW, how would you represent "mussels" - this came up for me in a
> problem I am working on .
>

Thanks for everyone's advice.

Here are my goals:

1) Create a unique id for each species which can point to
    entries in uBioNameBank, ITIS, and NCBI. This is for
    the concept of a species, for instance Anopheles barberi.
    The taxonomic names for mosquito and other insects
    are in to much flux for me to want to mark up photos etc.
    with only the current name.

2) Each instance of a species class would point to RDF
    documents containing information on that species including
    photo's, collection records, and other biological information

   For instance:
    Specieshas -> http://ex.org/ID3245_collections.rdf
    Specieshas -> http://ex.org/ID3245_standard_photos.rdf
    Specieshas -> http://ex.org/ID3245_hypothesized_distribution.rdf
    Specieshas -> http://ex.org/ID3245_host_species_list.rdf
    Specieshas -> http://ex.org/ID3245_predator_parasite_species_list.rdf
    Specieshas -> http://ex.org/ID3245_larval_habitat_requirements.rdf
    Specieshas -> http://ex.org/ID3245_list_of_WI_Counties_found_in.rdf


3) The list of attributes (Specieshas) would probably be the same for all
     the members of a given family, but not for all species, for instance
     all mosquitoes might have a common set of larval habitat descriptors,
     but that data might not included in a the schema for spiders, also
     the standard photo's would be of different structures.

The links between uBio NameBank and other uber banks to allow access
into and out of the datastore within the DC.coverage of Wisconsin.

Possible queries would be:

What culicidae (mosquitoes) have been collected in Dane County in April?

What predators of Anopheles barberi are found in Sauk County?

What is the range of dates that that predator species X has been
collected in Wisconsin?

What host plants for Danaus plexippus are found in Door County?

What viruses have are know to be found in the culicidae species that
have been collected in Barron County?

Has virus X ever been found in species Anopheles barberi in
Barron County?

Has any Anopheles barberi from Barron County been
tested for virus X?

My WGS-84 geo coordinates are xxxxx, xxxxxx what
mosquitoes are likely to
occur here?

I am less interested in taxonomic hierarch per se except
for retaining relationships for instance that the Diptera
(True Flies) of Dodge county would include culicidae (mosquitoes)

Thanks in Advance :-)

- Pete

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Organism Taxonomic Ontology

Alan Ruttenberg-2
Based on what you describe here, the taxonomy of levels (kingdom,
Phylum ... and downward) isn't really necessary for your queries. They
are necessary to help disambiguate names.

If you represented each specific organism as a class, with
superclasses for the genus, tribe, subfamily, etc, you would be able
to do all of these queries. You could include the level as an
annotation property. Here's what I mean (owl abstract syntax - rdfxml
available on request).

Namespace(ex = <http://example.com/#>)
Ontology(<http://example.com/#> Class(ex:Kingdom_Animalia partial
annotation(ex:level ex:Kingdom ))
Class(ex:Phylum_Arthropoda
  partial annotation(ex:level ex:Phylum)
  ex:Kingdom_Animalia )
Class(ex:Subphylum_Hexapoda
  partial annotation(ex:level ex:Subphylum)
  ex:Phylum_Arthropoda )
Class(ex:Class_Hexapoda
  partial annotation(ex:level ex:Class)
  ex:Subphylum_Hexapoda )
Class(ex:Subclass_Insecta
  partial annotation(ex:level ex:Subclass)
  ex:Class_Hexapoda )
Class(ex:Infraclass_Neoptera
  partial annotation(ex:level ex:Infraclass)
  ex:Subclass_Insecta )
Class(ex:Order_Diptera
  partial annotation(ex:level ex:Order)
  ex:Infraclass_Neoptera )
Class(ex:Suborder_Nematocera
  partial annotation(ex:level ex:Suborder)
  ex:Order_Diptera )
Class(ex:Infraorder_Culicomorpha
  partial annotation(ex:level ex:Infraorder)
  ex:Suborder_Nematocera )
Class(ex:Family_Culicidae
  partial annotation(ex:level ex:Family)
  ex:Infraorder_Culicomorpha )
Class(ex:Subfamily_Anophelinae
  partial annotation(ex:level ex:Subfamily)
  ex:Family_Culicidae )
Class(ex:Genus_Anopheles
  partial annotation(ex:level ex:Genus)
  ex:Subfamily_Anophelinae )
Class(ex:specific_epithet_barberi
  partial annotation(ex:level ex:specific_epithet)
  ex:Genus_Anopheles )
)

Then if you asked for the subclasses of ex:Family_Culicidae you get
ex:specific_epithet_barberi ex:Genus_Anopheles ex:Subfamily_Anophelinae

If you wanted to make annotation of species a little easier, given
that it sounds like most of your annotation will be there, then you
could make it an instance

Individual(<ex:specific_epithet_barberi>
  annotation(ex:level <ex:specific_epithet> )
  type(<ex:Genus_Anopheles>)
  value(ex:Specieshas <http://ex.org/ID3245_collections.rdf)
 )

Might want to give the different file types different property names.

This assumes that the combination of level and name is unique, e.g.
there isn't another organism with specific_epithet barberi. Don't know
if that holds.

-Alan


On 5/15/06, Pete DeVries <[hidden email]> wrote:

> Alan Ruttenberg <alanruttenberg <at> gmail.com> writes:
>
> >
> > Hi Pete.
> >
> > One thing you might think about is what inferences/queries do you want
> > to be able to do nand what constraints do you want automaticallly
> > checked? Given those, it would be easier to evaluate a particular
> > proposal.
> >
> > BTW, how would you represent "mussels" - this came up for me in a
> > problem I am working on .
> >
>
> Thanks for everyone's advice.
>
> Here are my goals:
>
> 1) Create a unique id for each species which can point to
>     entries in uBioNameBank, ITIS, and NCBI. This is for
>     the concept of a species, for instance Anopheles barberi.
>     The taxonomic names for mosquito and other insects
>     are in to much flux for me to want to mark up photos etc.
>     with only the current name.
>
> 2) Each instance of a species class would point to RDF
>     documents containing information on that species including
>     photo's, collection records, and other biological information
>
>    For instance:
>     Specieshas -> http://ex.org/ID3245_collections.rdf
>     Specieshas -> http://ex.org/ID3245_standard_photos.rdf
>     Specieshas -> http://ex.org/ID3245_hypothesized_distribution.rdf
>     Specieshas -> http://ex.org/ID3245_host_species_list.rdf
>     Specieshas -> http://ex.org/ID3245_predator_parasite_species_list.rdf
>     Specieshas -> http://ex.org/ID3245_larval_habitat_requirements.rdf
>     Specieshas -> http://ex.org/ID3245_list_of_WI_Counties_found_in.rdf
>
>
> 3) The list of attributes (Specieshas) would probably be the same for all
>      the members of a given family, but not for all species, for instance
>      all mosquitoes might have a common set of larval habitat descriptors,
>      but that data might not included in a the schema for spiders, also
>      the standard photo's would be of different structures.
>
> The links between uBio NameBank and other uber banks to allow access
> into and out of the datastore within the DC.coverage of Wisconsin.
>
> Possible queries would be:
>
> What culicidae (mosquitoes) have been collected in Dane County in April?
>
> What predators of Anopheles barberi are found in Sauk County?
>
> What is the range of dates that that predator species X has been
> collected in Wisconsin?
>
> What host plants for Danaus plexippus are found in Door County?
>
> What viruses have are know to be found in the culicidae species that
> have been collected in Barron County?
>
> Has virus X ever been found in species Anopheles barberi in
> Barron County?
>
> Has any Anopheles barberi from Barron County been
> tested for virus X?
>
> My WGS-84 geo coordinates are xxxxx, xxxxxx what
> mosquitoes are likely to
> occur here?
>
> I am less interested in taxonomic hierarch per se except
> for retaining relationships for instance that the Diptera
> (True Flies) of Dodge county would include culicidae (mosquitoes)
>
> Thanks in Advance :-)
>
> - Pete
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>
>
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Organism Taxonomic Ontology

Pete DeVries
Alan Ruttenberg <alanruttenberg <at> gmail.com> writes:

>
> Based on what you describe here, the taxonomy of levels (kingdom,
> Phylum ... and downward) isn't really necessary for your queries. They
> are necessary to help disambiguate names.
>
> If you represented each specific organism as a class, with
> superclasses for the genus, tribe, subfamily, etc, you would be able
> to do all of these queries. You could include the level as an
> annotation property. Here's what I mean (owl abstract syntax - rdfxml
> available on request).
....

Wow! thanks :-)

Sometimes there are species that have the same same
specific_epithet and they may be incompletely different
groups so it might be hard to detect them ahead of time.

I don't know if this helps but a scientific name is
traditionally the combination of genus and specific epithet
like Anopheles barberi often the author is also listed
so it would be Anopheles barberi Coquillett, and
sometimes Anopheles barberi Coquillett 1903 (the year
the species was described)

The other concept that might be good to try to represent is
that these taxonomic levels are based on what are believed
to be common shared characteristics since they evolved
from a common ancestor.

For instance true flies (Diptera) have two wings rather than
the four found in some other groups. What were the
second set of wings now exist as nub-like structures
called haltares that are thought to help with balancing.

Would this ontology be able to handle these kind of
shared characters? i.e. know that all diptera have two
wings?

 (at least like whales have hair like other mammals),

and if the specimen has six legs (insects) and two wings
it is very likely a dipteran?

Thanks!

- Pete

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Organism Taxonomic Ontology

Alan Ruttenberg-2

On May 16, 2006, at 9:22 PM, Pete DeVries wrote:

> Sometimes there are species that have the same same
> specific_epithet and they may be incompletely different
> groups so it might be hard to detect them ahead of time.
>
> I don't know if this helps but a scientific name is
> traditionally the combination of genus and specific epithet
> like Anopheles barberi often the author is also listed
> so it would be Anopheles barberi Coquillett, and
> sometimes Anopheles barberi Coquillett 1903 (the year
> the species was described)
Ok. So you can name it Anopheles_barberi_Coquillett. The main thing  
is that it is unique.

> The other concept that might be good to try to represent is
> that these taxonomic levels are based on what are believed
> to be common shared characteristics since they evolved
> from a common ancestor.
>
> For instance true flies (Diptera) have two wings rather than
> the four found in some other groups. What were the
> second set of wings now exist as nub-like structures
> called haltares that are thought to help with balancing.
>
> Would this ontology be able to handle these kind of
> shared characters? i.e. know that all diptera have two
> wings?
Yes. In the sample I've attached, I've defined two properties.  
hasNumberOfWings and hasNumberOfLegs.
The class example:Subphylum_Hexapoda has a "restriction" that says  
all instances have six legs. The class
Order_Diptera has a restriction that says all instances have 2 wings.

>  (at least like whales have hair like other mammals),
>
> and if the specimen has six legs (insects) and two wings
> it is very likely a dipteran?

You can't say "very likely". You can either say that if something has  
6 legs and 2 wings, then it definitely is dipteran (or is in the  
union of a specified listing of classes)

Or you can say suppose I have something which has 6 legs and 2 wings,  
what sorts of things might that be. I've done the latter in the  
example I've attached. To ask the question I've created a class  
specimen_class which is defined as exactly those things that have 6  
legs and 2 wings. If you use a reasoner to classify the ontology, and  
look at the inferred class tree, you will now find that Order_Diptera  
is a direct subclass of this class, since we know that all  
Order_Diptera have 2 wings (directly stated) and 6 legs (inherited  
from hexapoda). If there were other non diptera classes that also had  
these properties, then these would also be subclasses. But there  
isn't any way to define an ordering to decide which you prefer.

There are other ways of implementing this query, but this use of a  
"probe class" is easily demonstrated in protege.
Note that in my example I've represented "6" as the individual named  
"six" and "2" by the individual named "two". I've done this to get
around a limitation of the current way that protege connects to  
reasoners. This should be fixed in an upcoming version.

I've also attached the a second version that uses numbers directly.  
You'll need to use SWOOP to see the reasoner do the inference on this  
one.

Regards,
Alan


>
> Thanks!
>
> - Pete
>
> ----------------------------------------------------------------------
> ---
> To unsubscribe go to http://protege.stanford.edu/community/ 
> subscribe.html
>


6leg2wings.owl (3K) Download Attachment
6leg2wings-swoop.owl (3K) Download Attachment