Apollo Structured Vocabulary (Apollo-SV)

MB: Right now we represent 'purely intentional entities' as siblings to 'information content entities'. Given their nature I am not entirly sure that this is correct. It seems that both 'information content entity' and 'simulation' are subclass to 'purely intentional entity'.

MB: Here we run into the well know DL problem that we cannot specify that the infectious agent that is transmitted is of the same type that the infectious agent the population is susceptible to.

Comment from Mike (to do!): I checked because an infection and immunity census would involve two different census methods. One would be some kind of disease surveillance that counts infections. The other is some kind of immunity census process based on seroprevalence studies or vaccination registries.

For example, in a simulated volcano eruption, representations of lava, lava tubes, vents, etc. take the place of actual lava, lava tubes, vents, etc.

Even in the case of a simulated disaster scenario where actual ambulances, fire trucks, and people participate, representations of catastrophic events and injuries, etc. take the place of real ones.

Class was originally created to correspond with a complex type from Apollo XSD v3.1.0.

MB: In order to get the temporal occupation relations I added properties that I renamed with BFO2 URIs.

The duration of the period between when an organism becomes infected and when it becomes infectious.

MW: I know this causes problems with our set-theoretic definition. Unfortunately the compartmental models produce fractional counts at each of their time steps, including the final step.

Any ideas? I’ve considered and rejected allowing individual organisms to be fractional. Perhaps if we just replace the term “cardinality” in the definition with ‘computed size’

MB: This class is still awaiting its class restriction. It should be restricted as being about a population that has members that are eligible for a vaccination against the members that are eligible and consented the vaccination. There is a class "eligibility rule" in OBI (http://purl.obolibrary.org/obo/OBI_0500026), but it is restricted by definition to studies. I put in an issue in the OBI issue tracker. The consenting process and its specified outcome should be done in d-acts, but needs to be done properly.

Notes: 1. Apollo-SV does not distinguish between ‘vaccine efficacy’ and ‘vaccine effectiveness.’ 2. Apollo-SV also considers the use of ‘vaccine’ in both terms a misnomer

Motivating use case is Synthia synthetic population data.

Key motivating use case was Synthia synthetic population data.

a simulation that is about an organism

Key motivating use case was Synthia synthetic population data.

MB: For DL reasons the easiest way to formalize the definition was to assume only organism that are either host or not immune undergo antiviral treatment. Is that adequate? I assume this'll create problems…

MW: Since the estimates that modelers use are from real populations, I have two questions: 1. Should we distinguish between real and simulated population? 2. If so, do we want to represent a ‘measured’ antiviral compliance number that is a property of a real population and distinguish it from this number that we are using to study what might happen in a simulated population?

MB: This issue will need to be taken care of in the axiomatization of simulated population

MB: APOLLO_SV_00000053 and APOLLO_SV_00000058 have not been given a class restriction, since they are not expressible in BFO right now, due to their referring to "will". Both classes have been made disjoint in the OWL file, though.

MB: I am not 100% happy with the elucidation. Can't we just say that this is the full identifier identifying the simulator with both its name and its version number?

Motivating use case is to cover Synthia data, and possibly configuration/output of the MRSA simulator in the MIDAS community (CA-MRSA Repast).

This class will need more ontological reasoning, since the "unwilling" part of the definition is not representable in BFO to date.

MB: For all the antiviral-related classes I referred to ChEBI: Antiviral class. I wasn't able to use the plug-in though (there was a JAVA heap space error, probably due to the size of ChEBI). AND (more importantly) I did not re-use any of the ChEBI hierarchy which, even in the BFO version is a mess. Maybe we need to discuss this.

Key motivating use case was Synthia synthetic population data.

MB: The ontological status of simulations (Thoughts)

One property that seems to be common among simulations in comparison to many other fictions is the fact that the chronological order of the simulated purely intentional entities is not changed in the process of simulating.

While in many pieces of fiction the order in which the purely intentional entities are created can vary, this is not the case for simulations.

Class was originally created to correspond with an element from Apollo XSD v3.1.0.

simulation time step value simulation time step unit simulation run length

The time step value and unit are specified using a scalar value specification.

typically, the process endpoint is achievement of R0 < 1

The essential goal of an infectious disease control strategy is to protect the susceptible organisms in a population.

1. Apollo-SV does not distinguish between ‘vaccine efficacy’ and ‘vaccine effectiveness.’

2. Apollo-SV also considers the use of ‘vaccine’ in both terms a misnomer.

outer, gaseous layers of exoplanets are called extraterrestrial atmosphere.

infection is reproduction of a pathogen inside some tissue of a host. By associating infection with the classes latent period and infectious period, which are associated with specific pathogen species and host species, Apollo-SV defines infection further as reproduction of a particular pathogen taxon inside some tissue of a particular host taxon.

Cannot acquire from colonized host directly - contamination results from colonized or contaminated thing and new host picks it up from contaminated thing

a specification of a population that a control strategy applies to. It is intended to apply to a real population, so I don’t know whether you would consider it a simulated thing. I don’t.

Key motivating use case was Synthia synthetic population data.

per Mike Wagner, this is input

fraction of organisms (usually infected) that a surveillance system detects out of the total population

the fraction of some population that is targeted for a control strategy that accepts it. It can obviously also be interpreted as a probability that the average individual in the targeted population accepts/complies with the control strategy. It is not simulated

replaced with VO 'vaccine function'

it is the same thing as transmission coefficient

Comment from Mike (to do!): I checked because an infection and immunity census would involve two different census methods. One would be some kind of disease surveillance that counts infections. The other is some kind of immunity census process based on seroprevalence studies or vaccination registries.

need to change the text definition. need to change label to be more consistent with current Apollo terminology (possibly 'infectious disease scenario'.

A contaminated thing is an entity that contains a pathogen (either internal to itself or on its surface) that is not undergoing an infection.

Motivating use case is Synthia synthetic population data.

Motivating use case is Synthia synthetic population data.

Motivating use case is Synthia synthetic population data.

Motivating use case is Synthia synthetic population data.

Key motivating use case was Synthia synthetic population data.

An infectious disease scenario contains counts of infections of a particular pathogen in a particular host in a particular ecosystem at a particular point in time, minimally. Agent-based ecosystems represent each host individually.

The infectious disease scenario typically also includes a lot of additional data about the ecosystem, including data about things like transmission probabililites, control measures, the immunity or other resistance status of host individuals, and so on.

the preferred term is biological contamination to distinguish from things like radiological contamination, contamination of the Gulf of Mexico with oil, the contamination of air with pollution, etc.

This census should exclude infection, immunity, treatment, or contamination.

An instance of the an AgentLevel PopulationAndEnvironmentCensus type comprises an AgentLevel PopulationAndEnvironmentCensus Description, which describes the biotic and abiotic ecosystems elements that are counted by the census, and a URL where the census (the counts) can be retrieved.

At present, this class corresponds with an XSD type that contains authentication plus information needed to generate a Synthia synthetic population dataset including (1) geographical location and (2) a list of household characteristics.

The Apollo XSD allows us to specify counts of certain subpopulations and abiotic ecosystem elements for a compartmental model (vs. specifying each individual in the population as in agent model population and environment census).

A class representing an abstract type. This class has no attributes and programmers should use it to determine whether an instance should be saved or can be retrieved.

The infectious disease type in the XSD has a disease outcome element, which specifies whether someone became symptomatic, was hospitalized, etc. All of these outcomes require some process to occur. For sequalae of disease, it is that the realization of the disease is an etiological process that creates some other disorder. For example, fever is often occurs because of high levels of certain interleukins (which is a disorder).

So I am comfortable with disease outcome as a process. It is also consistent with the probabilitistic information that will go with it, as a future probability is almost always (if not always) the probability of some event occurring (e.g., coin turning up heads, raining, developing a disease),

Class was originally created to correspond with an element from Apollo XSD v3.1.0.

This is probably a 'process profile' according to BFO2

need to generate class restriction. Something like:

'is realized by' some (processual_entity and (has_participant some ('Anatomical structure' and ('is physical basis of' some 'protective resistance'))

we'd also like to say that the protective resistance disposition began to exist during the process.

can't recall whether we should use process or processual_entity for future BFO compatibility.

Some regard a clade to be a single branch on the tree of life. [http://en.wikipedia.org/wiki/Clade]