Perl 6 - the future is here, just unevenly distributed

IRC log for #bioclipse, 2015-09-01

| Channels | #bioclipse index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
03:16 cgfbee joined #bioclipse
04:15 egonw joined #bioclipse
04:46 zaetnick joined #bioclipse
05:25 sneumann_ joined #bioclipse
05:26 egonw joined #bioclipse
05:51 egonw @tell jonalv yes, the OPS 1.3 API has been phased out
05:51 zarah Consider it noted.
05:52 egonw @tell jonalv I am happy to discuss the plugin work on this channel
05:52 zarah Consider it noted.
05:52 egonw @tell @Gpox CDK 1.5.11 is out, with some work that probably will interest the Uppsala team: https://github.com/cdk/cdk/wiki/1.5.11-Release-Notes
05:52 zarah Consider it noted.
05:56 egonw @tell jonalv the BSL scripting book has some examples
05:56 zarah Consider it noted.
06:44 egonw joined #bioclipse
06:51 zaetnick joined #bioclipse
07:49 jonalv joined #bioclipse
08:07 egonw_ joined #bioclipse
08:46 Gpox joined #bioclipse
10:49 egonw_ joined #bioclipse
12:23 sneumann_ joined #bioclipse
12:33 sneumann_ joined #bioclipse
13:13 commitbot joined #bioclipse
13:13 commitbot [bioclipse.core] jonalv created jonalv-openPhacts-patches (+1 new commit): http://git.io/vGaMk
13:13 commitbot bioclipse.core/jonalv-openPhacts-patches 7d7ed69 jonalv: Created Bioclipse Search viewpart
13:13 commitbot left #bioclipse
13:19 zaetnick joined #bioclipse
13:52 commitbot joined #bioclipse
13:52 commitbot [bioclipse.core] jonalv pushed 1 new commit to jonalv-openPhacts-patches: http://git.io/vGajE
13:52 commitbot bioclipse.core/jonalv-openPhacts-patches 8a892b1 jonalv: Added search top menu and an icon
13:52 commitbot left #bioclipse
14:35 egonw_ jonalv: seen my new page? http://egonw.github.io/ (third moving image = Bioclipse-R)
14:38 jonalv that’s nice
15:08 egonw_ jonalv: btw, I so have two blog posts about ops4j, though at the R package wrapper, but that uses the same lib as the Bioclipse plugin
15:08 egonw_ http://chem-bla-ics.blogspot.nl/2013/12/open-phacts-from-r.html
15:08 egonw_ http://chem-bla-ics.blogspot.nl/2013/12/pharmacology-data-from-open-phacts-in-r.html
15:08 zarah egonw_'s link is also http://tinyurl.com/ob93aqu
15:09 jonalv egonw_: great
15:10 jonalv egonw_: I am messing around with Bioclipse search dialog
15:10 jonalv egonw_: a place to hook this in
15:10 egonw_ yes, saw the commits... why a 'core' search functionality?
15:11 jonalv egonw_: Neded a place to put the UI action
15:11 jonalv top menu seemed reasonable
15:11 jonalv also, that was were Ola ment to put it way back
15:12 jonalv egonw_: what I get back does not seem to be molecules
15:13 jonalv more precisely, openphacts.lookUpCompounds does not seem to return molecules
15:13 jonalv IMolecules even
15:14 jonalv CWResult is not something that plays nicely with BSL
15:14 jonalv There is some sort of mapping missmatch going on here
15:15 jonalv a CWResult seems to have a name and cwid, but what kind of data is it?
15:16 egonw_ CWResult is a hack by Ola and CW is dropped from the Open PHACTS project, so feel free to remove it...
15:16 egonw_ I had not done yet, because I had not talked with Ola about it
15:16 egonw_ CW is short for Concept Wiki
15:17 egonw_ and that service takes the role of IRS, Identity Resolution Service
15:17 egonw_ aka... name to identifier resolution
15:17 jonalv hm
15:18 jonalv I am desperately trying to understand what that has to do with what I am trying to do
15:18 jonalv clearly I need to learn more about the basics
15:18 egonw_ it's like this:
15:18 egonw_ most OPS API methods work with identifiers...
15:18 egonw_ there is one exception
15:19 egonw_ the API methods that take a human readable string
15:19 jonalv OPS?
15:19 egonw_ like "aspirin"
15:19 egonw_ OPS = Open PHACTS, duh :)
15:19 jonalv ah
15:19 egonw_ "aspirin" first needs conversion to an identifier (URI to be precise)
15:19 jonalv yea
15:20 egonw_ practically, in the current BC (Bioclipse ;) code, it's like:
15:20 jonalv I think, what the simple use case here was trying to do was just do a free text search and return any molecules found
15:20 egonw_ "aspirin" -> ConceptWikiIRI (aka CWResult) -> API calls
15:20 egonw_ yes
15:21 egonw_ "aspirin" -> URI (a conceptwiki URI) -> getAllPharmacology()
15:21 zaetnick joined #bioclipse
15:21 jonalv is it a bad idea to do this as one manager method?
15:21 egonw_ (internally, OPS will use the CW URI and map that to URIs used in other databases, such as the URI used by ChEMBL
15:21 egonw_ )
15:23 jonalv so that it becomes: public List<IMolecule> lookUpCompounds(String name)
15:23 egonw_ possibly...
15:23 egonw_ CW can "resolve" also proteins, diseases, etc, etc
15:23 egonw_ the API methods has a type...
15:24 jonalv also, I seem to always get 10 hits regardless of what I ask for. Weird...
15:24 jonalv 10 CWResults taht is
15:26 egonw_ default in either BC or the API
15:26 egonw_ there is an API option for that
15:27 jonalv hm, there is much work to be done here
15:27 jonalv clearly
15:27 egonw_ selecting another parameter value is not a lot of work, or?
15:27 jonalv It\s that I am thinking about it all wrong
15:28 jonalv I wanted to report number of hits to the user
15:28 jonalv so that the user could see that and make a judgement of what to do based on that
15:28 egonw_ ah, try the explorer...
15:29 egonw_ the IRS step is to select what "thing" the user is thinking about...
15:29 egonw_ that's not about the number of search results, but getting the right (tm) search result
15:29 egonw_ this "aspirin" and not that other entity also called "aspirin"
15:29 egonw_ (well, probably a bad example...)
15:30 jonalv hm, maybe
15:31 jonalv WE have been talking about general Search functionlaity in Bioclipse before so I thought this would be a good time to add a start for it
15:32 jonalv but the idea then was that one wanted to see the number of hits to know what to od with them (open them, save them directly to file or just ignore and search for something else instead(
15:33 egonw_ that's step too in Open PHACTS
15:33 jonalv that also ties well into that iterable thing we were talking about. Only this would be a interable thing with a known size
15:33 egonw_ you first need to make sure you know what the user *is* searching for
15:33 egonw_ 1. what type of entity
15:34 egonw_ 2. what type of result
15:34 jonalv hm, maybe that is what should be dealt with in the search dialog then?
15:35 egonw_ that makes sense
15:36 jonalv hm don’t really see how this can work. Do you ahve some specific use cases in mind?
15:38 jonalv I mean, in Bioclipse I guess searching for molecules is the one that makes sense for now?
15:48 egonw_ not sure...
15:48 egonw_ you may want to search for a target?
15:48 egonw_ Maris may be looking for kinases, or?
15:48 egonw_ I have to catch my train now...
15:48 egonw_ we'll continue talking!
15:48 egonw_ but later...
15:49 jonalv later o/
16:34 egonw_ joined #bioclipse
18:31 sneumann_ joined #bioclipse

| Channels | #bioclipse index | Today | | Search | Google Search | Plain-Text | summary