Better hyperlinking on pharmacology

[leeharland]

on http://jupiter.cs.man.ac.uk/trees/pharmacology?uri=http%3A%2F%2Fpurl.uniprot.org%2Fenzyme%2F4.-.-.-

"homo sapiens" is hyperlinked

• probably no need for now
• its hyperlinked to an EBI RDF page not a human-readable HTML page
• same for target name, should be linked to uniprot or something

essentially the explorer needs to put human hyperlinks there not the data ones supplied by the API

[ianwdunlop]

What is the relationship between http://rdf.ebi.ac.uk/resource/chembl/targetcomponent/CHEMBL_TC_702 and uniprot? Do the results from the cache have the uniprot link in them? What is the human readable version of http://rdf.ebi.ac.uk/resource/chembl/target/CHEMBL261?

[danidi]

The human readable version of the chembl target uri is https://www.ebi.ac.uk/chembl/target/inspect/CHEMBL261.

The first link you mention is a target component. In some cases (e.g. https://www.ebi.ac.uk/chembl/target/inspect/CHEMBL2095180) a chembl target is not only a single protein, but is e.g. a target family, which has several members. Each of these members is again a single protein (as far as I know), which should be available in uniprot. We don't have the uniprot links in the result, only conceptwiki.

[danidi]

In general, wouldn't it be better to use the hyperlinks in the result tables only for linking to other pages in the Explorer (e.g. Target information)? Links to ChEMBL, Uniprot,... should be done via Provenance in my opinion.

[Chris-Evelo]

This sounds to me like a user preference. Some users (including me) will want effective linkouts others will indeed expect links to stay inside the tool. Apart from what @danidi proposes you could 1) make them visually different or make the whole thing use configurable.