Replacing obsoleted GO terms

[sebastianduesing]

(Edited to add usage information for each of these terms.)

After running make imports for a routine import update (see #1805), I've found that GO has recently obsoleted some terms that were in use in the assay template, some of which have no equivalent in GO. We will need to decide on or create replacements for the following now-obsolete terms:

DNA methylation

GO:0006306

GO suggests "DNA-methyltransferase activity" (GO:0009008), "heterochromatin formation" (GO:0031507), or "DNA restriction-modification system" (GO:0009307) as replacements.

In OBI, DNA methylation is a biological process that is used as a target entity for the assay "DNA methylation profiling assay" (OBI:0000634), and it is used in an additional axiom for the device "Illumina methylation BeadChip" (OBI:0001870).

Regulation of DNA methylation

GO:0044030

GO suggests "enzyme regulator activity" (GO:0030234) or "epigenetic regulation of gene expression" (GO:0040029) as replacements. It's not clear to me whether either of these are sufficient for OBI's purposes.

In OBI, regulation of DNA methylation is a biological process that is used as a target entity for the following assays: "amplification of intermethylated sites assay" (OBI:0001685), "bisulfate sequencing assay" (OBI_0000748), "MeDIP-seq assay" (OBI:0000693), "methylation-sensitive restriction enzyme sequencing assay" (OBI:0001861), and "methylation-specific polymerase chain reaction assay" (OBI:0001684).

Histone modification

GO:0016570

GO does not suggest any replacements for this term.

In OBI, Histone modification is a biological process that is used as a target entity for two assays: "histone modification identification by ChIP-chip assay" (OBI:0002016) and "histone modification identification by ChIP-Seq assay" (OBI:0002017).

[bpeters42]

For histone modification (where there is no replacement), I am wondering how we are using that in OBI? As a material entity detected? If so, we might want to look outside of GO, and rather PRO ?

On Thu, Aug 1, 2024 at 12:56 PM Sebastian Duesing @.***> wrote:

After running make imports for a routine import update (see #1805 https://github.com/obi-ontology/obi/pull/1805), I've found that GO has recently obsoleted some terms that were in use in the assay template, some of which have no equivalent in GO. We will need to decide on or create replacements for the following now-obsolete terms: DNA methylation

GO:0006306 http://purl.obolibrary.org/obo/GO_0006306

GO suggests "DNA-methyltransferase activity" (GO:0009008 http://purl.obolibrary.org/obo/GO_0009008), "heterochromatin formation" (GO:0031507 http://purl.obolibrary.org/obo/GO_0031507), or "DNA restriction-modification system" (GO:0009307 http://purl.obolibrary.org/obo/GO_0009307) as replacements. Regulation of DNA methylation

GO:0044030 http://purl.obolibrary.org/obo/GO_0044030

GO suggests "enzyme regulator activity" (GO:0030234 http://purl.obolibrary.org/obo/GO_0030234) or "epigenetic regulation of gene expression" (GO:0040029 http://purl.obolibrary.org/obo/GO_0040029) as replacements. It's not clear to me whether either of these are sufficient for OBI's purposes. Histone modification

GO:0016570 http://purl.obolibrary.org/obo/GO_0016570

GO does not suggest any replacements for this term.

— Reply to this email directly, view it on GitHub https://github.com/obi-ontology/obi/issues/1806, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2ISKR52Y64YDYXP3ZY3ZPKHG7AVCNFSM6AAAAABL3HJSHWVHI2DSMVQWIX3LMV43ASLTON2WKOZSGQ2DGMRZGY2DONA . 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

[sebastianduesing]

@bpeters42 Histone modification is a biological process that is used as a target entity for two assays: "histone modification identification by ChIP-chip assay" (OBI:0002016) and "histone modification identification by ChIP-Seq assay" (OBI:0002017). I've added this information as well usage information for each of the other obsoleted terms to my original comment.

[bpeters42]

Let's do this on the call

On Fri, Aug 2, 2024 at 8:50 AM Sebastian Duesing @.***> wrote:

@bpeters42 https://github.com/bpeters42 Histone modification is a biological process that is used as a target entity for two assays: "histone modification identification by ChIP-chip assay" (OBI:0002016 http://purl.obolibrary.org/obo/OBI_0002016) and "histone modification identification by ChIP-Seq assay" (OBI:0002017 http://purl.obolibrary.org/obo/OBI_0002017). I've added this information as well usage information for each of the other obsoleted terms to my original comment.

— Reply to this email directly, view it on GitHub https://github.com/obi-ontology/obi/issues/1806#issuecomment-2265686897, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2IUVXS74LUFOAJKZ3PTZPOTEBAVCNFSM6AAAAABL3HJSHWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENRVGY4DMOBZG4 . You are receiving this because you were mentioned.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

[sebastianduesing]

Discussed on 2024-08-12 OBI dev call. Our plan is to make DNA methylation and regulation of DNA methylation into analyte assays, replacing the axioms that assert that they're target entities of various assays with axioms asserting those assays have ("DNA nucleotide" and "has part" some "methyl group"). For histone modification, we will be requesting that PRO creates classes for "histone" and "modified histone" for use as analytes in the assays using those terms.

[sebastianduesing]

(Edited to fix a link to the wrong pull request.)

An update on this: PRO has created 'modified histone', which we intend to use as an analyte in the assays that formerly referenced GO:'histone modification'. However, it won't be in the live release file until PRO creates the next release after their upcoming one, which was too far along in the development process for 'modified histone' to be added. It's not currently "visible" to Ontofox, and I believe it won't be until it's in the live release. I've reached out to ask how often PRO does releases, so I can estimate when we can expect to see 'modified histone' in the live release file of PRO, and I'll follow up here once I know that.

As far as the effort to refresh OBI's imports goes (see #1814), we've got a couple options on how to proceed:

• Wait to finalize that PR until PRO:'modified histone' is live.
• Finalize that PR in its current form, in which the two assays that used 'histone modification' have each lost the axiom referencing that term, then repair those two in a later PR after we can import PRO:'modified histone'.

[bpeters42]

I would go with the second, and 'fix later'

Another thing this reminds me off: For the longest time we wanted to have placeholder terms; imagine a 'RT:23432' term for 'Requested term', which has a placement in the ontology and a link to an issue tracker, and a note that this will eventually be replaced. That is what we have been doing for our IEDB work, but it seems like it could be a general workflow and not be very scary.

On Tue, Aug 20, 2024 at 11:04 AM Sebastian Duesing @.***> wrote:

An update on this: PRO has created 'modified histone http://purl.obolibrary.org/obo/PR_000085778', which we intend to use as an analyte in the assays that formerly referenced GO:'histone modification'. However, it won't be in the live release file until PRO creates the next release after their upcoming one, which was too far along in the development process for 'modified histone' to be added. It's not currently "visible" to Ontofox, and I believe it won't be until it's in the live release. I've reached out to ask how often PRO does releases, so I can estimate when we can expect to see 'modified histone' in the live release file of PRO, and I'll follow up here once I know that.

As far as the effort to refresh OBI's imports goes (see #1805 https://github.com/obi-ontology/obi/pull/1805), we've got a couple options on how to proceed:

• Wait to finalize that PR until PRO:'modified histone' is live.
• Finalize that PR in its current form, in which the two assays that used 'histone modification' have each lost the axiom referencing that term, then repair those two in a later PR after we can import PRO:'modified histone'.

— Reply to this email directly, view it on GitHub https://github.com/obi-ontology/obi/issues/1806#issuecomment-2299440840, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADJX2ISYQBSAQ5O55IENHCTZSOAJ3AVCNFSM6AAAAABL3HJSHWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOJZGQ2DAOBUGA . You are receiving this because you were mentioned.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

[sebastianduesing]

I'm happy to go with the "fix later" option as well—both of the two assays that would lose an axiom temporarily have 0 usages within OBI, so it doesn't seem likely to cause any downstream issues within OBI to leave them as is until the PRO term is live. As for the placeholder terms thing, I see what you mean—that reminds me of the HCC placeholder IRIs. I agree that there's a use-case for OBI/OBO more generally.

[sebastianduesing]

(Edited because I used the wrong term name.)

Ah, I just saw Darren's response—looks like it'll be 6 months or so until 'modified histone' is in the live release file, and thus until we can import it with Ontofox. The "fix later" option seems like the best way to go, given that timeline.

[sebastianduesing]

A follow-up on this: I've experimented with manually adding 'modified histone' to the PRO import file by converting PRO's OBO format stanza to OWL and adding that manually to src/ontology/OntoFox_imports/PR_imports.owl. It's possible to do so, but not ideal; it would mean that we'd have to be very careful running make imports until the next PR release, as refreshing that import would delete that term from the file. It also wouldn't save much work later on, as I'd still want to rerun make src/ontology/OntoFox_imports/PR_imports.owl after the next PRO release, because there's a chance that the OntoFox import of that term would shuffle some lines around a bit from my converted manual import.

I've never worked with thesrc/ontology/external-byhand.owl file before, so I'm not certain that this is actually what this file is for, but if that file is a source for manual, non-Ontofox imports, we could put 'modified histone' there for the time being, which would solve the problem of being careful with make imports, but would also require a fix later on after the release.

Either way, some work will be required at the time of release. I'm still perfectly happy to leave those two assays without an axiom for the time being—the logical cost of doing so doesn't seem terribly high, it saves us (me) from doing some extra work now, and it requires the same or less work later than the other options.