Open egonw opened 6 years ago
@mkutmon, is there something ready to test now?
ping @JonathanMELIUS @mkutmon, what is the state of this issue? If I run the curl command, it seems it is still not returning someting, but the linked lucene-indexer issue is marked closed?
@egonw I don't know how the WikiPathways webservice works in details. But the issue seems to definitely be the BridgeDb version used there, an old datasource.txt file.
The pathway is indexed and the webservice uses the metabolite database since we get some results to those calls: https://webservice.wikipathways.org/getXrefList?pwId=WP3620&code=Ca https://webservice.wikipathways.org/getXrefList?pwId=WP3620&code=Ch
@ariutta Can you identify the source of BridgeDb information for the latest Annotation Panel (and datanode table linkouts) on pathway pages?
We should make a plan to consolidate the sources of BridgeDb info (including the datasources.txt). There were as many as 3 at one point. Not sure how many there are now. Ideally, there would be one.
@JonathanMELIUS I followed all the webservice stack and didn't find anything that could create issues with the call. I think the problem might be that the service at http://vm4.wikipathways.org:8983/ is outdated. That's the server that needs to be updated. I briefly checked this in the afternoon with Jonathan but I didn't realized that this was the server that was being called.
Can you identify the source of BridgeDb information for the latest Annotation Panel (and datanode table linkouts) on pathway pages?
@AlexanderPico the Annotation Panel and datanode table linkouts are both using the XrefPanel.php extension.
To address this issue more generally, how about we always make the latest datasources.txt and datasources_headers.txt available from bridgedb.org? We'll need to be sure bridgedb.org has CORS enabled for those files, so that JS resources on wikipathways.org can access them.
@AlexanderPico, yes, we should aim at consolidating... for now, updating what we have sounds useful to me too.
@ariutta, I filed a feature request for the BridgeDb webservice to provide more info about data sources: https://github.com/bridgedb/BridgeDbWebservice/issues/34
@egonw @AlexanderPico @ariutta,
@mkutmon show us that it is not a code issue but an outdated metabolite database issue: https://webservice.wikipathways.org/getXrefList?pwId=WP7&code=Cks
The webservice returns KNApSAcK (Cks) ids for the WP7.
And on https://www.wikipathways.org/index.php/Pathway:WP3620, the Cks ids do map in the "External references" table, because it gets the information from webservice.bridgedb.org
So basically, we just need to update the databases on the index server (vm4.wikipathways.org).
I think to be more consistent all the ids mapping in WikiPathways servers ( index, webservice, pvjs, external references table, ... ) should use the same instance of BridgeDb in vm4.wikipathways.org
Then webservice.bridgedb.org would be free from WikiPathways servers traffic ??
I am trying to use Wikipathways as a data source of our (@afukushima and me) Metabolite Set Enrichment Analysis software. I need your cooperation for that. Would it be difficult to update vm4?
I thought it was updated... @nunogit @JonathanMELIUS did your account access work?
@AlexanderPico It seems that the update has no effect about this issue yet. https://webservice.wikipathways.org/getXrefList?pwId=WP3620&code=Cks
@egonw Do you know where the fix to this issue lies? Should the updated databases address this?
@AlexanderPico no, I was not able to access the server.
As reported here and with a test:
This is what the Webservice UI gives:
@JonathanMELIUS, do you have ideas about this? Is the webservice using the latest BridgeDb 2.2 and metabolite mapping data?