Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Egon Willighagen
    @egonw
    ok, that works for me too now... so that was a temp glitch .... closing it...
    ah, right... one sec...
    Ian Dunlop
    @ianwdunlop
    It was possibly a timeout for the explorer issue. The first time round the browser gets fed up waiting for virtuso to do something. But it does eventually and has cached it.
    Egon Willighagen
    @egonw
    query to confirm the first first :) ->
    this has two interactions, with these _about IRIs:
    and
    from the last bit (which is copied from the GPML, the WikiPathways data format) that the interaction has the same ID (= the same)
    and you can see here that two RDF versions are loaded...
    that I deduce from the fact that pathway WP2261 has two revisions
    and a single WPRDF download always only has the latest revision of a pathway
    Ian Dunlop
    @ianwdunlop
    Haven't got any spare time to look into this right now so I'll add it to the ever increasing list of things to do when I clone myself :(
    Egon Willighagen
    @egonw
    yes, I was chatting with Nick about available resources...
    last year it was already hard to use in the programming course...
    but right now, we cannot link any pathway information to pharmacology anymore...
    I think we need to sit down what we want Open PHACTS to be, with the very limited resource we have...
    Ian Dunlop
    @ianwdunlop
    BTW I just did a quick test using a different entity and it also appears in multiple 'latest' revisions.
    Ian Dunlop
    @ianwdunlop
    If you try the following query on the beta sparql endpoint you get 2 revisions:
    PREFIX dc: <http://purl.org/dc/elements/1.1/> PREFIX dcterms: <http://purl.org/dc/terms/> SELECT DISTINCT ?rev WHERE { GRAPH <http://www.wikipathways.org> { ?rev dc:identifier <http://identifiers.org/wikipathways/WP2261> . ?item dcterms:isPartOf ?rev . } }
    On a fresh triple store with the latest wp loaded ie /home/ubuntu/d/era/a/downloads/wikipathways/wikipathways-20170310-rdf-wp.zip you get 1 revision.
    Egon Willighagen
    @egonw
    exactly
    good point about VOS deleting of data... honestly, I do not have the loading processes clear, so did not think of that as possible cause
    Ian Dunlop
    @ianwdunlop
    I guess that 20171010 is the latest version (and not the 20170310 that I did the quick test on).
    Egon Willighagen
    @egonw
    @randykerber, @ianwdunlop, Chris mentioned a v3 beta... but I do not remember that... do you know more about it?
    Ian, yes, that number is a date...
    WikiPathways have monthly updates...
    on the 10th of each month
    Ian Dunlop
    @ianwdunlop
    v3 It was a prototype for a sustainable OP platform. More a demonstrator for a future direction than a replacement. The prototype contained chembl & uniprot with compound and target info. No wp data yet.
    Egon Willighagen
    @egonw
    ok, clear
    and interesting, btw...
    does v3 come with a Docker Composer (or whatever it is called) that we could try on the OpenRiskNet cloud?
    to provide feedback?
    so, we established in our group that at least the IMS Docker is useful, also with other data, and ready for repurposing...
    Ian Dunlop
    @ianwdunlop
    It's all dockerized. Not sure whether its allowed out in the wild yet. Nick/Bryn/OPF may help you out there.
    Egon Willighagen
    @egonw
    a IMS-enriched-SPARQL-endpoint would be next for us, I think
    Ian Dunlop
    @ianwdunlop
    Using sparql directly via the linksets or calling out to an IMS within the sparql?
    Anwyay. I see you mentioned that using HGNC ids doesn't work. There doesn't seem to be any hgnc to other dataset linksets currently. Were there before?
    Egon Willighagen
    @egonw
    [ims/sparql] fair point... sorry, I'm particularly uncapable of thinking today...
    I meant, whatever the component it is that takes an initial SPARQL query, runs it through the QueryExpander (that uses the IMS), and then sends it to the SPARQL of the cache...
    [hgnc] I am not sure, but we could easily provide the data... just not sure what we should be providing...
    Egon Willighagen
    @egonw
    @ianwdunlop, @randykerber, Jonathan sent me the HGNC linksets yesterday... please let me know if I can forward those for (re)loading
    Ian Dunlop
    @ianwdunlop
    @Egon. Feel free to forward them. Then we can figure out the logistics for reloading. I get a cold sweat just thinking about that :(
    Egon Willighagen
    @egonw
    :)
    ok, forwarding that email now then.
    thanks
    Egon Willighagen
    @egonw
    plz let me know if you did not receive it yet
    Ian Dunlop
    @ianwdunlop
    Can't see anything in the inbox
    Ian Dunlop
    @ianwdunlop
    @egonw got the email. Now to find some time to do something about it.....