Hi Andy (cc-ed the public jalview-dev list for posterity)
We are wrestling with Jalview’s Ensembl abstraction and wondering whether the struggle is worth it. We currently provide two Ensembl ‘data sources’ - Ensembl and EnsemblGenomes - and are also having to do some additional work to resolve the EnsemblFungi/etc Division based cross-references that Uniprot serve.
From a UX perspective it is (IMHO) preferable to have a single ‘Ensembl’ datasource for resolution of any identifier, but because there are separate endpoints, under the hood we still need to distinguish identifiers for divisions held at EnsemblGenomes.
I seem to remember you mentioning that there was a plan that once ES-Ensembl was up and running, EnsemblGenomes would probably disappear (or just proxy to the same endpoint as Ensembl). Is that still going to be the case ? If yes, and it is going to happen in the next year, then that might help us decide the most appropriate technical solution here.
Jim.
PS. It just occured to me that the only reason Uniprot holds division-specific Ensembl identifiers is that it allows the Uniprot web site to provide web-links to the different division front-ends. I guess that reason is good enough for Jalview to do the same - but the multiple end-point discussion still stands 
···
-------------------------------------------------------------------
Dr JB Procter, Jalview Coordinator, The Barton Group
Division of Computational Biology, School of Life Sciences
University of Dundee, Dundee DD1 5EH, UK.
+44 1382 388734 | [www.jalview.org](http://www.jalview.org) | [www.compbio.dundee.ac.uk](http://www.compbio.dundee.ac.uk)
Awesome - thanks Andy !
Looking forward to the shiny new unified endpoint 
… j.
Hi Jim and Jalview developers
To clarify a few things here that I hope will help your decision process
- The medium term idea is to provide Ensembl/EG data through a single unified interface. We are planning to finish this work on merging the two REST servers into one URL by Q1 2019 (release 96). We don't envisage developer having to change anything on their side (though the websites will not be merging)
- There is a pan Ensembl (including EG) id lookup redirect for the website from [http://www.ensemblgenomes.org/id/](http://www.ensemblgenomes.org/id/){id} e.g. [http://www.ensemblgenomes.org/id/ENSG00000139618](http://www.ensemblgenomes.org/id/ENSG00000139618) and [http://www.ensemblgenomes.org/id/AT3G52430](http://www.ensemblgenomes.org/id/AT3G52430) redirect to their respective website
- ES-Ensembl is unlikely to be a production system until 2020 so holding on for that may not be the right thing to do
Finally I believe the logic you have about UniProt holding division-specific identifiers is for URL reconstruction. We can provide ID redirection, as shown above, but any genome location links need to the which site it's hosted on.
Hope this helps,
Andy
Hi Andy (cc-ed the public jalview-dev list for posterity)
We are wrestling with Jalview's Ensembl abstraction and wondering whether the struggle is worth it. We currently provide two Ensembl 'data sources' - Ensembl and EnsemblGenomes - and are also having to do some additional work to resolve the EnsemblFungi/etc Division based cross-references that Uniprot serve.
>From a UX perspective it is (IMHO) preferable to have a single 'Ensembl' datasource for resolution of any identifier, but because there are separate endpoints, under the hood we still need to distinguish identifiers for divisions held at EnsemblGenomes.
I seem to remember you mentioning that there was a plan that once ES-Ensembl was up and running, EnsemblGenomes would probably disappear (or just proxy to the same endpoint as Ensembl). Is that still going to be the case ? If yes, and it is going to happen in the next year, then that might help us decide the most appropriate technical solution here.
Jim.
PS. It just occured to me that the only reason Uniprot holds division-specific Ensembl identifiers is that it allows the Uniprot web site to provide web-links to the different division front-ends. I guess that reason is good enough for Jalview to do the same - but the multiple end-point discussion still stands ;)
--
-------------------------------------------------------------------
Dr JB Procter, Jalview Coordinator, The Barton Group
Division of Computational Biology, School of Life Sciences
University of Dundee, Dundee DD1 5EH, UK.
+44 1382 388734 |
[www.jalview.org](http://www.jalview.org) | [www.compbio.dundee.ac.uk](http://www.compbio.dundee.ac.uk)
The University of Dundee is a registered Scottish Charity, No: SC015096
The University of Dundee is a registered Scottish Charity, No: SC015096
···
On 25/09/2018 12:14, Andy Yates wrote:
--
-------------------------------------------------------------------
Dr JB Procter, Jalview Coordinator, The Barton Group
Division of Computational Biology, School of Life Sciences
University of Dundee, Dundee DD1 5EH, UK.
+44 1382 388734 | [www.jalview.org](http://www.jalview.org) | [www.compbio.dundee.ac.uk](http://www.compbio.dundee.ac.uk)
FYI - I’ve moved this discussion back to the issue that started this thread:
https://issues.jalview.org/browse/JAL-2755
···
On 25/09/2018 12:18, James Procter wrote:
On 25/09/2018 12:14, Andy Yates wrote:
--
-------------------------------------------------------------------
Dr JB Procter, Jalview Coordinator, The Barton Group
Division of Computational Biology, School of Life Sciences
University of Dundee, Dundee DD1 5EH, UK.
+44 1382 388734 | [www.jalview.org](http://www.jalview.org) | [www.compbio.dundee.ac.uk](http://www.compbio.dundee.ac.uk)
--
-------------------------------------------------------------------
Dr JB Procter, Jalview Coordinator, The Barton Group
Division of Computational Biology, School of Life Sciences
University of Dundee, Dundee DD1 5EH, UK.
+44 1382 388734 | [www.jalview.org](http://www.jalview.org) | [www.compbio.dundee.ac.uk](http://www.compbio.dundee.ac.uk)