UNIMOD in OLS

[ypriverol]

@edeutsch @germa

As we discussed in Beijing the Unimod Library has been added to OLS http://www.ebi.ac.uk/ols/ontologies/unimod This is one of the steps to move forward UNIMOD as a central place for Mass spectrometry Modifications. The guys from OLS would like that we improve the obo representation of unimod. But this is up to us.

Regards Yasset

[edeutsch]

This is great, thanks. But I do not understand what you/they mean by “improve the obo representation”. It would seem to be entirely sufficient to me. I use it all the time. It includes allowed site information and such. Maybe there are some things that are visible on the web site that are not encoded in obo, but I have not found any that I needed. So it would seem to me that what they currently produce is entirely adequate. But maybe I’m missing something…

Eric

From: Yasset Perez-Riverol [mailto:notifications@github.com] Sent: Monday, May 08, 2017 7:17 AM To: HUPO-PSI/psi-ms-CV Cc: Eric Deutsch; Mention Subject: [HUPO-PSI/psi-ms-CV] UNIMOD in OLS (#6)

@edeutsch https://github.com/edeutsch @germa https://github.com/germa

As we discussed in Beijing the Unimod Library has been added to OLS http://www.ebi.ac.uk/ols/ontologies/unimod This is one of the steps to move forward UNIMOD as a central place for Mass spectrometry Modifications. The guys from OLS would like that we improve the obo representation of unimod. But this is up to us.

Regards Yasset

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/HUPO-PSI/psi-ms-CV/issues/6, or mute the thread https://github.com/notifications/unsubscribe-auth/AMHnhkk-xFZmYSlbvc2irpJ8h20_gfc2ks5r3yPWgaJpZM4NUAs0 .

[ypriverol]

Hi @edeutsch : If you see this example: http://www.ebi.ac.uk/ols/ontologies/go/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FGO_0098743

compare with any of the Unimod examples, you will notice the difference. For example the description of the term is missing in UNIMOD, the xref is different also from other databases which actually are the properties of each modification.

I will contact the guys from OLS @simonjupp that really knows from which fields in the obo file these xref are populated.

Regards Yasset

[edeutsch]

Hmm, for example a UNIMOD term:

http://www.ebi.ac.uk/ols/ontologies/unimod/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FUNIMOD_4

I see a def and a synonym all fine.

But the xrefs do seem to be a strangely rendered I suppose..

Maybe the heart of the problem is that they use the xref construction to encode a lot of attributes of the term, which is not really intended by the designers of obo. Maybe there’s a better way.

Note that we do the same thing in the MS ontology for encoding our value-types:

http://www.ebi.ac.uk/ols/ontologies/ms/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FMS_1000858

From: Yasset Perez-Riverol [mailto:notifications@github.com] Sent: Monday, May 08, 2017 1:21 PM To: HUPO-PSI/psi-ms-CV Cc: Eric Deutsch; Mention Subject: Re: [HUPO-PSI/psi-ms-CV] UNIMOD in OLS (#6)

Hi @edeutsch https://github.com/edeutsch : If you see this example: http://www.ebi.ac.uk/ols/ontologies/go/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FGO_0098743

compare with any of the Unimod examples, you will notice the difference. For example the description of the term is missing in UNIMOD, the xref is different also from other databases which actually are the properties of each modification.

I will contact the guys from OLS @simonjupp https://github.com/simonjupp that really knows from which fields in the obo file these xref are populated.

Regards Yasset

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/HUPO-PSI/psi-ms-CV/issues/6#issuecomment-299979272, or mute the thread https://github.com/notifications/unsubscribe-auth/AMHnhhqqh6iTwWR2kAPWtCJccJk5-xV6ks5r33kcgaJpZM4NUAs0 .