looking for everyone to comment on where they see exposures fitting in overall

[diatomsRcool]

As you can see, Nico and I are starting to migrate ECTO to odk and adding patterns that are a bit different from what was there originally. We are trying to be in alignment with PECO and ExO. There is still a lot of work to do. Feel free to take a look at the patterns in the dosdp-patterns directory (the others are just in reserve and may not be used). I would love some feedback.

[bpeters42]

It would be fantastic if this alignment could include the alignment with our work on Immune Exposure, which as discussed are essentially a subset of what you are covering. But we have different axes of classification of 'how the exposure comes about' and 'evidence for the exposure', which I think will be useful. Specifically: 'Administered exposures', (e.g. medications taken) vs. 'Environmental exposures' (e.g. living in an Malaria endemic area), or 'Accidental exposures' (needleprick accident in a hospital). Along with that comes the evidence for the exposure, which is particular important for the 'Environmental exposure' ones, (e.g. lab tests such as serology giving confidence that exposure actually happened).

Our preference has been to postcompose exposure terms. It looks like you want to create them all up front. Both is fine, but we need to coordinate on the patterns.

[bpeters42]

So ideally we would distinguish between describing the shared part of all exposures 'organism X is exposed to Y' (as an aside: I really hope we only care about organisms? not e.g. how my car paint exposed to sunlight withers?) And further subclass this to 'where is the organisms exposed' or in ECTO: where is the 'receptor'? And provide independent models on how intended exposures are administered, and how evidence on environmental exposures is collected.

I hope this is making sense even though I am writing in a rush. I am trying to say we should look to what is 'mind-independent' (quoting Chris Mungall) about exposures first, model that, and then have a separate set of patterns on modeling administrations and evidence.

[diatomsRcool]

1. Yes, we are only talking about organisms here
2. ECTO is meant to be agnostic to organism or "receptor"

The patterns only contain the stressor, route, and medium. Is it possible that your work on this could add the receptor, 'how the exposure comes about', and 'evidence for the exposure'? Could we just declare these bits out of scope for ECTO? Although I see we may need a pattern for exposure to a biological entity or something like that to cover what you need.

[bpeters42]

Last time I looked at ECTO, it was a very heterogeneous mixture of stuff, and briefly looking at the current ecto-full.owl file, seeing 'injection' and 'surgical manipulation' very much seems to be in the domain of 'how an exposure comes about'. Maybe I am misunderstanding, and you could point me at what file I should be looking at, and what classes I should take serious, and which ones are just along for the ride (as I understand this is a mapping exercise to some extend)?

In general, we had pointed out early on that we are very interested to be involved in this, and have already done considerable work in the domain, so it would be great if we can integrate work on this more closely.

• Bjoern

On Mon, Aug 12, 2019 at 7:44 AM diatomsRcool notifications@github.com wrote:

1. Yes, we are only talking about organisms here
2. ECTO is meant to be agnostic to organism or "receptor"

The patterns only contain the stressor, route, and medium. Is it possible that your work on this could add the receptor, 'how the exposure comes about', and 'evidence for the exposure'? Could we just declare these bits out of scope for ECTO? Although I see we may need a pattern for exposure to a biological entity or something like that to cover what you need.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/EnvironmentOntology/environmental-exposure-ontology/issues/39?email_source=notifications&email_token=ADJX2IWOAWHUVMEFHCSZFPTQEFZOVA5CNFSM4IKXHKK2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4CYICQ#issuecomment-520455178, or mute the thread https://github.com/notifications/unsubscribe-auth/ADJX2ISPLSRPMHXQFZTRPT3QEFZOVANCNFSM4IKXHKKQ .

-- Bjoern Peters Professor La Jolla Institute for Allergy and Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

[matentzn]

This sounds great @bpeters42 ! We are currently in the process of redoing some of ECTO, but many of the classes you saw last time will be added back in over the course of the next two months. I would be great to have everything 100% aligned with your immune exposures. None of our patterns are baked in stone yet. Check for example this one: exposure_to_environmentalprocess_route.

environmental process: ENVO:02500000
route: ExO:0000055

'exposure event' and ('has exposure stimulus' some %s) and ('has exposure route' some %s)

As you can see, we have a stimulus (stressor) here which is an environmental process (earthquake) and a route which should come from EXO (could be something like physical contact, injection etc, although this needs to be extended for this here pattern).

If we didnt mention a route:

environmental process: ENVO:02500000

'exposure event' and ('has exposure stimulus' some %s)

The class would simply classify above the former one. Can you share some examples of exposures that you are working on?

[diatomsRcool]

@bpeters42 yeah, a lot of that stuff is legacy and we're still sorting it out. I anticipate using MAXO for the medical stuff. Leigh and I will be working on that. The ones you should take serious are in the dosdp-patterns folder. Sorry its a mess, but we're in the middle of a major update. I think we really want to make sure we align with immune exposure when talking about exposure to organisms and exposure to infections. I could use some guidance there.

[cmungall]

The surgical stuff is a distraction. That could all be moved out.

On Mon, Aug 12, 2019 at 8:24 PM bpeters42 notifications@github.com wrote:

Last time I looked at ECTO, it was a very heterogeneous mixture of stuff, and briefly looking at the current ecto-full.owl file, seeing 'injection' and 'surgical manipulation' very much seems to be in the domain of 'how an exposure comes about'. Maybe I am misunderstanding, and you could point me at what file I should be looking at, and what classes I should take serious, and which ones are just along for the ride (as I understand this is a mapping exercise to some extend)?

In general, we had pointed out early on that we are very interested to be involved in this, and have already done considerable work in the domain, so it would be great if we can integrate work on this more closely.

• Bjoern

On Mon, Aug 12, 2019 at 7:44 AM diatomsRcool notifications@github.com wrote:

1. Yes, we are only talking about organisms here
2. ECTO is meant to be agnostic to organism or "receptor"

The patterns only contain the stressor, route, and medium. Is it possible that your work on this could add the receptor, 'how the exposure comes about', and 'evidence for the exposure'? Could we just declare these bits out of scope for ECTO? Although I see we may need a pattern for exposure to a biological entity or something like that to cover what you need.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/EnvironmentOntology/environmental-exposure-ontology/issues/39?email_source=notifications&email_token=ADJX2IWOAWHUVMEFHCSZFPTQEFZOVA5CNFSM4IKXHKK2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4CYICQ#issuecomment-520455178 , or mute the thread < https://github.com/notifications/unsubscribe-auth/ADJX2ISPLSRPMHXQFZTRPT3QEFZOVANCNFSM4IKXHKKQ

.

-- Bjoern Peters Professor La Jolla Institute for Allergy and Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/EnvironmentOntology/environmental-exposure-ontology/issues/39?email_source=notifications&email_token=AAAMMOIXFMVRSXVYF4KX3R3QEISNXA5CNFSM4IKXHKK2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4ENV3Y#issuecomment-520674031, or mute the thread https://github.com/notifications/unsubscribe-auth/AAAMMOOJKN4EFNCNMWGUJLLQEISNXANCNFSM4IKXHKKQ .

[diatomsRcool]

@bpeters42 as far as the immune exposures go, I was imagining that ECTO would contain the actual exposure event and your work would use that ECTO class to model all the other stuff that is important. If memory serves, that's where we left it after our phone call. However, I do need help sorting these types of exposures. I will start a new issue. We can continue taking general comments here.