Integrating MRIO with OBI

[adbartni]

Hello, and thank you for taking the time to meet with us on the call before the holidays! The presentation from that call containing the background of MRI and MRIO as well as some use cases can be found here. We just had a couple questions before we proceed with a formal PR.

Since we think MRI acquisition and analysis is an important piece of the biomedical investigation process, we believe that some of our terms from MRIO have a place directly in OBI. The main term that we would like to include in OBI is our image data set (MRIO:0000623) class, which we feel best captures the metadata and additional information found in MRI, as well as other biomedical imaging techniques and digital image file formats. It should be noted that image data set isn't intended to replace image (http://purl.obolibrary.org/obo/IAO_0000101) from IAO; rather, we are developing a data visualization child class in MRIO to handle the relationship between image and image data set.

However, this has led to our developing a duplicate term for magnetic resonance imaging assay (introduced in #1291) with a disjoint definition, which we call MR imaging assay (MRIO:0000623). Our term produces an image data set as a specified output, while magnetic resonance imaging assay (http://purl.obolibrary.org/obo/OBI_0002985) from OBI inherits the has_specified_output some image axiom from its parent imaging assay (http://purl.obolibrary.org/obo/OBI_0000185) class.

We would like to use the preexisting OBI terms as much as possible and see two possible directions, if the OBI team wishes to proceed with incorporating our proposed terms.

1. Request to open magnetic resonance imaging assay for editing to bring in some of our axioms, OR
2. Request to open imaging assay for editing so that we can include an axiom stating that it and its subclasses produce an image data set (provided image data set is approved and included, of course). I'm not sure how the best way to do this would be, since some subclasses of imaging assay may need to produce an image and not an image data set.

I'm attaching a full list of terms we'd like to contribute to OBI and the definitions in this csv (OBI_terms.csv) if someone would like to view them. It seems that it's probably best to break the list into a few separate parent classes (i.e. image data set, MR imaging assay, and MR image data set analysis) and submit PRs for each. However, I've included the full list if it makes more sense to submit just one PR for all terms.

We would be happy to join another Monday call to discuss next steps!

More info can be found in the MRIO repo (in the obi branch). MRIO-base.owl can be imported into OBI or src/ontology/MRIO_for_OBI.owl can be opened on its own to view our terms.

[bpeters42]

Thank you for your detailed ticket Alexander,

• nothing is 'closed' in OBI, and your group has clearly thought more about MRI terms than we have, so we should use that to improve the current definitions. That being said, as current terms are in use across a number of projects, we are committed to not changing their meaning in a significant way. I think we should be doing both 1 + 2 based on this.
• I am not clear if 'image data set' is supposed to be a subclass of 'image'? The IAO definition of image isn't the best, but I think the intent is for it to capture what you describe as 'image data set' ?
• I like your suggestion of splitting it into 3 PRs

On Mon, Jan 10, 2022 at 4:05 PM Alexander Bartnik @.***> wrote:

Hello, and thank you for taking the time to meet with us on the call before the holidays! The presentation from that call containing the background of MRI and MRIO as well as some use cases can be found here https://github.com/Buffalo-Ontology-Group/MRI_Ontology/raw/obi/docs/MRIO_OBI.pptx. We just had a couple questions before we proceed with a formal PR.

Since we think MRI acquisition and analysis is an important piece of the biomedical investigation process, we believe that some of our terms from MRIO have a place directly in OBI. The main term that we would like to include in OBI is our image data set (MRIO:0000623) class, which we feel best captures the metadata and additional information found in MRI, as well as other biomedical imaging techniques and digital image file formats. It should be noted that image data set isn't intended to replace image ( http://purl.obolibrary.org/obo/IAO_0000101) from IAO; rather, we are developing a data visualization child class in MRIO to handle the relationship between image and image data set.

However, this has led to our developing a duplicate term for magnetic resonance imaging assay (introduced in #1291 https://github.com/obi-ontology/obi/pull/1291) with a disjoint definition, which we call MR imaging assay (MRIO:0000623). Our term produces an image data set as a specified output, while magnetic resonance imaging assay (http://purl.obolibrary.org/obo/OBI_0002985) from OBI inherits the has_specified_output some image axiom from its parent imaging assay (http://purl.obolibrary.org/obo/OBI_0000185) class.

We would like to use the preexisting OBI terms as much as possible and see two possible directions, if the OBI team wishes to proceed with incorporating our proposed terms.

1. Request to open magnetic resonance imaging assay for editing to bring in some of our axioms, OR
2. Request to open imaging assay for editing so that we can include an axiom stating that it and its subclasses produce an image data set (provided image data set is approved and included, of course). I'm not sure how the best way to do this would be, since some subclasses of imaging assay may need to produce an image and not an image data set.

I'm attaching a full list of terms we'd like to contribute to OBI and the definitions in this csv (OBI_terms.csv https://github.com/obi-ontology/obi/files/7842029/OBI_terms.csv) if someone would like to view them. It seems that it's probably best to break the list into a few separate parent classes (i.e. image data set, MR imaging assay, and MR image data set analysis) and submit PRs for each. However, I've included the full list if it makes more sense to submit just one PR for all terms.

We would be happy to join another Monday call to discuss next steps!

More info can be found in the MRIO repo https://github.com/Buffalo-Ontology-Group/MRI_Ontology (in the obi branch). MRIO-base.owl can be imported into OBI or src/ontology/MRIO_for_OBI.owl can be opened on its own to view our terms.

— Reply to this email directly, view it on GitHub https://github.com/obi-ontology/obi/issues/1481, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2IUJOL3QVPRGORE4WALUVNX23ANCNFSM5LU6K5FA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you are subscribed to this thread.Message ID: @.***>

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

[adbartni]

After discussing with the MRIO team, we'd like to proceed with 3 separate PRs as described above. We don't want to change the meaning or usage of terms that are already in OBI, and we'd really like to use the pre-existing magnetic resonance imaging assay term. With that in mind, our first PR will be the introduction of the image data set class, since we don't feel that image fully captures all of the metadata and time series information that we talk about with MRI images. We'd also like to introduce a planned process that defines how an image may be produced by an image data set, as well as expand the definition of imaging assay to include image data set as an output without changing the rest of the definition.

All together, we plan on contributing 21 terms. Is there an IRI range we should use for our terms in the PR?

[bpeters42]

Hi Alexander,

Sorry for the late reply. I am hosting the call tomorrow, and wanted to go over your PRs, but I just saw that you were asking for IRI ranges. We don't reserve IRI ranges, but rather go through a 'get the next available' process. The fact that I couldn't immediately find the documentation on how to do that on the OBI website confirms that it is not just you. If you are available tomorrow, it would be great if you can join the OBI call. If not, we will still cover at least the IRI issue and get back to you.

(And I should say, from my point of view, the terms you are proposing sound very good).

• Bjoern

On Sat, Feb 12, 2022 at 1:14 PM Alexander Bartnik @.***> wrote:

After discussing with the MRIO team, we'd like to proceed with 3 separate PRs as described above. We don't want to change the meaning or usage of terms that are already in OBI, and we'd really like to use the pre-existing magnetic resonance imaging assay term. With that in mind, our first PR will be the introduction of the image data set class, since we don't feel that image fully captures all of the metadata and time series information that we talk about with MRI images. We'd also like to introduce a planned process that defines how an image may be produced by an image data set, as well as expand the definition of imaging assay to include image data set as an output without changing the rest of the definition.

All together, we plan on contributing 21 terms. Is there an IRI range we should use for our terms in the PR?

— Reply to this email directly, view it on GitHub https://github.com/obi-ontology/obi/issues/1481#issuecomment-1037484577, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2IQEP2Z4EOYMBG3GQGLU23ESFANCNFSM5LU6K5FA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you commented.Message ID: @.***>

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

[jamesaoverton]

For editing OBI, the place to start is the README for this repository: https://github.com/obi-ontology/obi

There is a "Making Changes" section toward the end: https://github.com/obi-ontology/obi#making-changes

It links to the OBI Term ID Reservations Google Sheet: https://docs.google.com/spreadsheets/d/1tpDrSiO1DlEqkvZjrDSJrMm7OvH9GletljaR-SDeMTI/edit#gid=224833299

Striking the right balance with documentation is hard. Let us know if we need changes.

[bpeters42]

I think the level of granularity in the documentation is fine. My lack of ability to find the right section was because I thought the 'Editing' section would include the 'Making changes' section.

I don't have an immediately better solution. And I think going through stuff like this on the call every once in a while is a positive, to make sure more of the OBI people are aware how things are done and where the information is.

On Mon, Feb 28, 2022 at 9:18 AM James A. Overton @.***> wrote:

For editing OBI, the place to start is the README for this repository: https://github.com/obi-ontology/obi

There is a "Making Changes" section toward the end: https://github.com/obi-ontology/obi#making-changes

It links to the OBI Term ID Reservations Google Sheet: https://docs.google.com/spreadsheets/d/1tpDrSiO1DlEqkvZjrDSJrMm7OvH9GletljaR-SDeMTI/edit#gid=224833299

Striking the right balance with documentation is hard. Let us know if we need changes.

— Reply to this email directly, view it on GitHub https://github.com/obi-ontology/obi/issues/1481#issuecomment-1054481327, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2IX4PLZIEXQ6IFZ5VZTU5OU4VANCNFSM5LU6K5FA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you commented.Message ID: @.***>

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

[adbartni]

Agreed, I just missed that section of the documentation. I'll proceed with the ID reservations and PR once I do a little refactoring per Alan Ruttenberg.

[adbartni]

Do I need to be added as a contributor to the repo? I'm getting an access denied when I try to push my changes:

git push --set-upstream origin mrio-image-data-set-1481
ERROR: Permission to obi-ontology/obi.git denied to adbartni.
fatal: Could not read from remote repository.

Please make sure you have the correct access rights
and the repository exists.

I've been having other unrelated problems with pushes since setting my account up for 2FA last August, so if it's not an access setting on the repo I can try with my work account instead of my .edu account.

[jamesaoverton]

I invited you to the OBI Developers team, which should provide the right permissions.