Update of EC mappings

[mcourtot]

from @hattrill:

Not sure what your update pipeline for EC mappings is in the GO….. but just to let you know that EC:2.1.1.125 and EC:2.1.1.126 have been deprecated.

[tberardini]

@mcourtot , @ukemi, @paolaroncaglia , @dosumis , @cmungall : What does deprecation of an EC number mean? Is there an established procedure on how to handle terms that have a deprecated EC number in either a definition xref or a synonym xref? If we don't have one, let's develop an SOP. When we've done that, I will implement for this set of EC numbers.

There are several terms that contain these EC numbers.

[ukemi]

In the past we have just removed them.

[hdrabkin]

Will we get alerts emailed to us so we can remove them from the ontology? I assume "depreciation" means its gone and needs to be removed so we don't get EC2GO that don't really map to anything?

Harold

From: Tanya Berardini notifications@github.com<mailto:notifications@github.com> Reply-To: geneontology/go-ontology reply@reply.github.com<mailto:reply@reply.github.com> Date: Thursday, November 17, 2016 at 4:32 PM To: geneontology/go-ontology go-ontology@noreply.github.com<mailto:go-ontology@noreply.github.com> Subject: Re: [geneontology/go-ontology] Update of EC mappings (#12802)

@mcourtothttps://github.com/mcourtot , @ukemihttps://github.com/ukemi, @paolaroncagliahttps://github.com/paolaroncaglia , @dosumishttps://github.com/dosumis , @cmungallhttps://github.com/cmungall : What does deprecation of an EC number mean? Is there an established procedure on how to handle terms that have a deprecated EC number in either a definition xref or a synonym xref? If we don't have one, let's develop an SOP. When we've done that, I will implement for this set of EC numbers.

There are several terms that contain these EC numbers.

You are receiving this because you are subscribed to this thread.

Reply to this email directly, view it on GitHubhttps://github.com/geneontology/go-ontology/issues/12802#issuecomment-261376378, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AJ9NkAAuOHw4xqy3_IEyrEtpYiOVpLSQks5q_MgGgaJpZM4K1OLA.

The information in this email, including attachments, may be confidential and is intended solely for the addressee(s). If you believe you received this email by mistake, please notify the sender by return email as soon as possible.

[tberardini]

In the past we have just removed them.

"them" = the dxrefs or the terms themselves?

[hdrabkin]

The xrefs: this is like what Chris just did for the EC. X.Y.- xrefs.

Then the EC2GO file will not have a mapping to the GO term

Harold

From: Tanya Berardini notifications@github.com<mailto:notifications@github.com> Reply-To: geneontology/go-ontology reply@reply.github.com<mailto:reply@reply.github.com> Date: Thursday, November 17, 2016 at 4:43 PM To: geneontology/go-ontology go-ontology@noreply.github.com<mailto:go-ontology@noreply.github.com> Cc: Harold Drabkin Harold.Drabkin@jax.org<mailto:Harold.Drabkin@jax.org>, Comment comment@noreply.github.com<mailto:comment@noreply.github.com> Subject: Re: [geneontology/go-ontology] Update of EC mappings (#12802)

In the past we have just removed them.

"them" = the dxrefs or the terms themselves?

You are receiving this because you commented.

Reply to this email directly, view it on GitHubhttps://github.com/geneontology/go-ontology/issues/12802#issuecomment-261379055, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AJ9NkFnQfOPbLwJgxP9LmtggaHFrALWxks5q_MqKgaJpZM4K1OLA.

The information in this email, including attachments, may be confidential and is intended solely for the addressee(s). If you believe you received this email by mistake, please notify the sender by return email as soon as possible.

[ukemi]

Sorry. I was on my way out. If the activity is thought to no longer exist then we should obsolete the term. I think I remember doing that in the past. If for some reason we still think the activity is a valid molecular function, then we should just remove the cross-reference. You are correct though. If everyone agrees that this is the correct procedure, then we should write up an SOP and put it in the editor's guide. We can either make that as part of this ticket or you can make an new one and assign it to me.

[tberardini]

For these particular EC numbers:

GO:0008469, histone-arginine N-methyltransferase activity (and children)

EC 2.1.1.125 - histone-arginine N-methyltransferase deleted. Now covered by EC 2.1.1.319, type I protein arginine methyltransferase, EC 2.1.1.320, type II protein arginine methyltransferase, EC 2.1.1.321, type III protein arginine methyltransferase and EC 2.1.1.322, type IV protein arginine methyltransferase

GO:001627, [myelin basic protein]-arginine N-methyltransferase activity

EC 2.1.1.126 - [myelin basic protein]-arginine N-methyltransferase http://www.brenda-enzymes.org/enzyme.php?ecno=2.1.1.125 deleted. Now covered by EC 2.1.1.319, type I protein arginine methyltransferase, EC 2.1.1.320, type II protein arginine methyltransferase, EC 2.1.1.321, type III protein arginine methyltransferase and EC 2.1.1.322, type IV protein arginine methyltransferase

There are four different mechanisms for protein arginine methyltransferase. Looks like we'll need to (1) create the four new MF terms (check other existing children of 'protein-arginine N-methyltransferase activity' to be sure) (2) obsolete the two existing MF terms with the new ones as consider terms

What to do with the children of GO:0008469, histone-arginine N-methyltransferase activity? (see pic) They've got residue specificity. Obsolete and recapture the detail in AE? Retain but as direct children of 'protein-arginine N-methyltransferase activity'?

[mcourtot]

I had tagged as 'infrastructure' because I wanted to ask (which I subsequently forgot to do ;)) whether there could be an option to deal with those like we do imports? So at least when they get depreciated we could get an alert or remove them automatically?

[tberardini]

@hdrabkin @mcourtot @ukemi @paolaroncaglia

Reiterating the plan above: (1) create the four new MF terms (check other existing children of 'protein-arginine N-methyltransferase activity' to be sure) (2) obsolete the two existing MF terms with the new ones as consider terms

This is the part I need help with - please comment.

What to do with the children of GO:0008469, histone-arginine N-methyltransferase activity? (see pic) They've got residue specificity. Obsolete and recapture the detail in AE? Retain but as direct children of 'protein-arginine N-methyltransferase activity'?

[ukemi]

Coming back to this now and having second thoughts. Although it would be ideal to capture the histone terms with annotation extensions or in Noctua, I'm not sure we are ready for that yet. I remember that there was some discussion when we created these terms and the curator argument was similar to the one for the retention of the modified protein terms, they are too important not to have as instantiated terms in the ontology. So I would vote that for now we keep them, perhaps under GO:0016274, which we would retain.

So this might be a case where we keep the terms for the time being and we simply remove that EC xrefs as Harold originally suggested.

[tberardini]

I could keep all the existing terms, remove the deprecated EC xrefs AND add these four new MF terms:

EC 2.1.1.319, type I protein arginine methyltransferase EC 2.1.1.320, type II protein arginine methyltransferase EC 2.1.1.321, type III protein arginine methyltransferase EC 2.1.1.322, type IV protein arginine methyltransferase

How does that sound?

[ukemi]

Works for me.

[paolaroncaglia]

Me too, thanks @tberardini

[hdrabkin]

Yes;

[tberardini]

Turns out that all the individual types of protein arginine methyltransferases were already in GO.

I updated EC crossreferences, added synonyms and comments where appropriate, and removed the deprecated EC xrefs from the terms that had them.