The use of Disease Ontology definitions for diseases

[PhiBabs935]

In the IDO paper, we said that we should use Disease Ontology definitions for diseases. As noted in the paper, the problem here is that DO is notoriously messy in terms of following the OGMS model.

Werner Ceusters has suggested that instead of trying to get DO to curate their definitions, we should instead create new terms/definitions that are faithful to OGMS. Then we can defend in the paper why we couldn't reuse the DO terms (or will not continue to use them in the case of IDO extensions where we did so already). Hopefully this will force DO to clean up their disease terms. I am inclined to follow W's lead here, but we should probably all chat about this in a call at some point (along with various other issues).

I think that we should consider going ahead with definitions according to the template provided by IDOSA: Staphylococcus aureus infectious disease =def “Infectious disease that has a staphylococcus aureus infectious disorder as its material basis.” W told me that he thinks this follows OGMS just fine.

Thus, AIDS would be something like: "Infectious disease that has an HIV infectious disorder as its material basis" Influenza would be something like: "Infectious disease that has an influenza virus infectious disorder as its material basis" and so on....

[linikujp]

It seems that DO has programmatically provided the textual definitions. Also, DO uses an IDO relation has_material_basis_in to link the relation between infectious disease and its infectious agent, such as:

Class: viral infectious disease Term IRI: http://purl.obolibrary.org/obo/DOID_934 Definition: A disease by infectious agent that results in infection, has_material_basis_in Viruses

Unfortunately, the IDO hasn't give a definition for has_material_basis_in. It seems like the range for has_material_basis_in is an infection, but not the infectious agent. If this is the case, we would need to define it in IDO.

Is there a direct relation in IDO from an infectious disease to the agent (virus or bacteria) that causes the disease? Because comparing to infectious disorder, the agent is more important data to capture in science or clinical practice.

[PhiBabs935]

_has_material_basisin will eventually be replaced with a newer version of this relation from BFO-ISO (once it is published). You can find how the relations are defined here: https://buffalo.app.box.com/v/bfo-iso-owl-cl/file/576585613537

IDO Core is pathogen-neutral, so it itself doesn't have the direct connection to the agents for specific diseases. But this is something that can easily be done in IDO extensions, following the model provided John Beverly in IDO Virus and IDO COVID-19

The idea is that the causative pathogen is a part of the infectious disorder, so insofar that the disease is directly connected to (has material basis in) the infectious disorder, it is thereby directly connected to the pathogens that make up that infectious disorder.

In IDO Virus, John has given the term virus disorder (which is a subclass of infectious disorder) the following OWL definition:

infectious disorder and (has part some virus) and (inverse (has material basis in) some (realized in some viral disease course))

So, insofar that a viral disease is connected to an infectious disorder, it is logically connected to the viruses that make up that disorder.

Thus, consider John Beverly's IDO COVID 19 (you could ask John whether he could send you the owl file)

COVID-19 has a very complex OWL definition, but as pertains to your question, one conjunct of the OWL definition is this:

(has material basis in some SARS-CoV-2 disorder)

In turn, the term SARS-CoV-2 disorder has the OWL definition: coronavirus disorder and (has part some SARS-CoV-2) and ((inverse (has material basis in) some (realized in some COVID-19 disease course))

Thus, insofar that COVID-19 is directly connected to a SARS-CoV-2 disorder, it is automatically directly connected to the SARS-CoV-2 viruses that make up that disorder

[johnbeve]

I'm going to close this since it looks like we're going in the OGMS direction, and since and Shane's helpfully spelled out the recipe for logically linking disorder to causative agent.