Camelia, the Perl 6 bug

IRC log for #bioclipse, 2013-04-19

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

All times shown according to UTC.

Time Nick Message
04:28 hybridivy joined #bioclipse
04:58 hybridivy joined #bioclipse
05:27 sneumann joined #bioclipse
06:17 zaetnick joined #bioclipse
06:32 konditorn joined #bioclipse
07:01 hybridivy joined #bioclipse
07:05 egonw joined #bioclipse
07:15 hybridivy joined #bioclipse
07:38 Gpox joined #bioclipse
07:41 hybridivy joined #bioclipse
07:45 egonw joined #bioclipse
07:54 olass joined #bioclipse
08:58 jonalv joined #bioclipse
09:59 hybridivy joined #bioclipse
10:05 sravanthisinha joined #bioclipse
10:05 sravanthisinha joined #bioclipse
10:05 sravanthisinha joined #bioclipse
10:15 hybridivy joined #bioclipse
10:28 hybridivy left #bioclipse
10:29 hybridivy joined #bioclipse
10:29 hybridivy hi egonw and jonalv
10:29 hybridivy i have some quires with the project idea
10:30 egonw shoot
10:30 hybridivy queries*
10:30 hybridivy the 2d visulaisation which we are expecting is it same as its is done in cytoscape ?
10:31 hybridivy using vizmapper there?
10:32 hybridivy i mean when we load a certain pathway we can have any visual style - sbgn, or any visual style?
10:33 egonw IDEA 15, right?
10:33 hybridivy yes
10:33 egonw I was not aware PV supported more than one viz style :)
10:33 egonw I have ever only used the default
10:34 hybridivy pv doesnot cytoscape does
10:34 hybridivy but pv supports mim, sbgn
10:34 hybridivy so we can intepret gpml into diff
10:35 egonw does this mim and sbgn have other file formats?
10:35 egonw and what do you mean with "into diff" >
10:35 egonw and what do you mean with "into diff" ?
10:36 hybridivy i meant into differnet formats like sbgn and mim
10:37 egonw OK, yes, certainly possible!
10:37 egonw let me look up how to associate file formats with "Editors"
10:38 hybridivy yes they are in .sbgn and .miml formats
10:38 egonw see: https://github.com/bioclipse/bioclipse​.cheminformatics/blob/master/plugins/n​et.bioclipse.cdk.jchempaint/plugin.xml
10:38 egonw line 26ff shows the binding
10:38 egonw the formats are defined as (for example): net.bioclipse.contenttypes.mdlMolFile3D
10:39 egonw in this case, the definition is in a separate file...
10:39 egonw let me look it up...
10:39 egonw https://github.com/bioclipse/bioclipse.​cheminformatics/blob/master/plugins/net​.bioclipse.chemoinformatics/plugin.xml
10:39 egonw line 160 for this format
10:40 egonw so, yes, I like the idea of supporting these three formats as project goal
10:40 egonw hybridivy: btw, AFAIK jonalv is having lunch right now
10:40 hybridivy ok
10:41 hybridivy right now as i see the cml files have atoms and they are shown in the jmol editor or cml editor
10:42 hybridivy its either atoms or molucules
10:42 egonw yeah, CML is very flexible
10:42 egonw it can contain a lot of different things
10:42 egonw depending on the content, a specific editor is fired up
10:42 hybridivy but as of gpml its all genes and metabolite
10:43 egonw there is only one flavor of GPML
10:43 egonw so, a one to one relation between .gpml files and the PV-based editor
10:43 egonw that is:
10:43 hybridivy so do i take them as  basic atomic particles and represent in 2d and 3d
10:43 egonw we also use the PV editor to open .gpml files
10:43 hybridivy ?
10:43 egonw sorry... you lost me now
10:44 egonw what do you try to do?
10:44 egonw I would not open a GPML file in a Jmol or JChempaint editor...
10:44 hybridivy yes
10:45 hybridivy it should be opened in its own kind off editor right ?
10:46 hybridivy sorry.. carry on with what you were saying. i just messed up in 1D
10:47 egonw well, I was done with my answer
10:48 egonw what do you mean with 1D in your last comment?
10:48 olass joined #bioclipse
10:49 hybridivy i menat i was going on and on in only my one direction
10:50 hybridivy ok shall we start again with my doubts?
10:50 hybridivy :)
10:50 egonw yes
10:51 hybridivy ok ad jonalv said  Bioclipse can open 3d coordinates of molecules which are stored in files  for 3d structural information
10:52 hybridivy so i should take the coordinates from gpml file for each entity and represent it - that is 3d visualisation part right?
10:52 egonw ah, yes, but GPML does not have 3D coordinates for moleucles
10:52 egonw ah, yes, but GPML does not have 3D coordinates for molecules
10:52 egonw or 2D
10:53 egonw or even a connection table
10:53 egonw PathVisio can get the last for some metabolites from BridgeDB
10:53 egonw hybridivy: have you looked up ISelectionEvents ?
10:54 egonw check this page: http://wiki.eclipse.org/FAQ_How_do_I_make_a_view​_respond_to_selection_changes_in_another_view%3F
10:54 zarah egonw's link is also http://tinyurl.com/bosmua9
10:54 egonw and use Google for further info
10:54 egonw but the idea is this:
10:54 egonw GPML gets opened in the PathVisio editor
10:54 egonw this editor will send around ISelectionEvent's
10:55 egonw where the event will contain an IMolecule with a SMILES string as provided by BridgeDb
10:55 egonw start with that
10:59 hybridivy ok the SMILES will be taken by the jmol or jchempaint and show it
10:59 hybridivy right?
10:59 egonw right
10:59 hybridivy great
10:59 egonw for that, read the ISelectionEvent design in Eclipse
10:59 hybridivy i got now a clear picture of it
10:59 hybridivy thanks egonw
11:00 hybridivy :)
11:01 hybridivy ok sure i will read up, the technology part could be learnt up
11:02 egonw yes, but it is important to get the design clear in your proposal
11:02 egonw yes, but it is important to get the conceptual design clear in your proposal
11:03 hybridivy right egonw , i will work on it
11:04 egonw did you decide yet which project you would prefer to do?
11:04 egonw and thus focus your energy on?
11:04 hybridivy yes this idea and the publication refernce one for sure
11:06 olass joined #bioclipse
11:06 egonw OK
11:08 hybridivy and depend on the community , which they would approve
11:08 egonw yeah, that will only become more clear when we enter the submit phase
11:09 egonw hybridivy: just remember that many mentors already know you, and what you did in the past year
11:09 egonw that is an advantage you have over other applicants
11:10 egonw but a good proposal is really important to convince also the mentors that do not know you
11:10 egonw (and those that do know you, in that you are up for the job)
11:11 hybridivy right  egonw, i will submit the proposal soon, so that i can improve on the proposal right from start if anything required
11:12 hybridivy thanks egonw for the support , right from beginning
11:12 hybridivy yes im not related to biinformatics , sometimes it doesnt gwt to me quickly
11:13 hybridivy bioinformatics*
11:13 hybridivy but  i try to manage as quick as possible
11:13 egonw well, GSoC is about learning new things :)
11:15 hybridivy and the bioclipse idea is exciting and at the same time challenging too.
11:15 egonw quite so
11:17 hybridivy ok i will now work on the things you suggested me to about the project idea
11:17 hybridivy and get back to you , if any queries arise
11:17 hybridivy :)
11:17 egonw OK
11:17 egonw focus on a list of goals and the matching time line
11:18 egonw the list can literally be a list
11:18 egonw certainly in the first iteration
11:18 egonw after you have a list ready, then you can start writing up what is needed to make those goals reality
11:18 egonw that will be the body of your proposal
11:19 egonw but get that list clear first
11:19 egonw that will give you the right focus in the writing
11:19 egonw (and it can change later, if it turns out unrealistic)
11:21 hybridivy egonw ++
11:22 egonw maybe you can draft up the proposal in a Google Drive doc?
11:23 egonw shared with me and Jonathan?
11:23 egonw and make sure to check with GSoC what the template is
11:23 hybridivy right egonw, first i will concentrate on the list and make a draft and share it with both of you
11:24 egonw yes
11:24 egonw that list can be in that doc too :)
11:27 hybridivy this is jonalv emailid: jonathan.alvarsson@gmail.com right?
11:27 egonw yes, and since this channel is logged we do not normally post email addresses here :)
11:27 egonw but I think the logger cuts it out, so you should be good :)
11:28 egonw http://colabti.org/irclogger/irclogge​r_log/bioclipse?date=2013-04-19#l155
11:28 egonw "this is jonalv emailid: [email@hidden.address] right?"
11:28 hybridivy oops sorry.. :)
11:28 hybridivy good part of irc log :)
11:29 egonw indeed :)
11:41 hybridivy egonw and jonalv , i have shared the initial doc on drive
11:41 egonw ok, thanx
11:42 egonw I have to leave in about 15 minutes
11:42 hybridivy ok
11:42 egonw so will read it later
11:42 hybridivy have a good day egonw :)
11:48 hybridivy joined #bioclipse
12:02 hybridivy joined #bioclipse
12:02 konditorn joined #bioclipse
12:17 jonalv hybridivy: I am back :) Just backlogged
12:29 hybridivy I have shared a basic doc with you on the project idea as suggested by egonw
12:29 jonalv cool
12:29 hybridivy i  will work on the idea and get back to you jonalv
12:29 hybridivy :)
12:30 jonalv hybridivy: One thing which I wonder is: Can you get 3d coordinates from PathVisio and if not, where can you get 3d coordinates?
12:32 sravanthisinha joined #bioclipse
12:35 hybridivy joined #bioclipse
12:35 jonalv hybridivy: One thing which I wonder is: Can you get 3d coordinates from PathVisio and if not, where can you get 3d coordinates?
12:37 hybridivy jonalv: im checking on it
12:37 jonalv okey
12:39 hybridivy jonalv:  <Graphics CenterX="484.4071875788865" CenterY="686.0936537954046" Width="60.0" Height="20.0" ZOrder="-18" FontWeight="Bold" FontSize="12" Valign="Middle"/>
12:39 hybridivy i guess the zorder does make it 3d
12:40 jonalv hybridivy: What is it you have pasted? :)
12:41 hybridivy i have pasted it from gmpl file, which is the format pathvisio primarily uses
12:41 jonalv okey
12:42 jonalv And what is this Graphics element? I am not sure I understand what the represents
12:42 jonalv *it represents
12:43 hybridivy every pathway elemnt has the graphics part of it, through which the elemnt is placed while viewing
12:44 jonalv ah but that is the graphics for the pathway. I meant 3d structures for the metabolites or proteins
12:45 hybridivy jonalv :http://developers.pathvisi​o.org/wiki/EverythingGpml
12:47 hybridivy jonalv: i think the 3d co-ordinatesof the metabolite or protein is same as the graphics part of gpml of that pathway element (datanode - that could be metabolite/gene)
12:48 jonalv hybridivy: hm, I have no idea.
12:49 hybridivy jonalv: for eg : in a .cml file at present has  for eg<atom id="a1" elementType="C" x3="3.249900" y3="-0.070900" z3="-0.279200"/>
12:49 jonalv yea, that looks like cml :)
12:49 hybridivy jonalv: that forms the 3d co-ordinates right?
12:49 jonalv yes
12:49 hybridivy yes
12:50 mgerlich joined #bioclipse
12:50 sneumann joined #bioclipse
12:51 cruttkie joined #bioclipse
12:51 hybridivy so similarly we can  get 3d co-ordinates  from the  gpml file http://svn.bigcat.unimaas.nl/pvplu​gins/Pathway-Validator/trunk/test-​files/gpml/WP528_41084_pass_0.gpml
12:51 zarah hybridivy's link is also http://tinyurl.com/d829qf6
12:52 jonalv hybridivy: where in that file do you see atom 3d coordinate?
12:54 jonalv I believe for example if you look at: http://pastebin.com/TLiJdxvA one has to use the CAS ID and look that one up somewhere
12:56 hybridivy jonalv: i had the same doubt whr does i have atom 3d in file of gpml, then egon suggested that when i open a gpml in pathvisio editor,the editor will send around ISelectionEvent's
12:56 hybridivy where the event will contain an IMolecule with a SMILES string as provided by BridgeDb
12:56 hybridivy the SMILES will be taken by the jmol or jchempaint and show it
12:57 jonalv hybridivy: yea, from SMILES it is possible to generate a 2D image and show it using jchempaint. That is a good first step
12:58 jonalv hybridivy: As for 3D coordinates. Generating those are usually a bit more complicated and it is probably a better idea to look them up I would guess.
13:00 hybridivy jonalv: right, any suggestions on how to proceed about the 3d co-ordinates?
13:02 jonalv hybridivy: I feel we might have to ask egonw if maybe he had some thoughts there which I am missing. Probably it is possible to look up and dowload 3d coordinates though the web somehow. If they in fact are not available. I just don't know. 2D structures from the SMILES string is however a good start :)
13:03 hybridivy jonalv: ok great :)
13:21 hybridivy jonalv: http://www.chembiogrid.org/cheminfo/smi23d/ where they claim SMILES strings to 3D.
13:21 hybridivy looking onto it
13:31 jonalv hybridivy: yea, perhaps generation is an alternative. We have Nalloon in Bioclipse for 3d generation as well. But the question is what is best...
13:31 jonalv *Balloon
13:32 hybridivy jonalv: great, ok i will check out with balloon too
13:33 egonw joined #bioclipse
13:35 jonalv lo egonw just the person to ask questions :)
13:41 hybridivy hi egonw
13:41 hybridivy i and jonalv were discussing about , how to get 3d coordinates from smiles
13:42 hybridivy egonw: then jonalv  suggested we could use Ballon in bioclipse to do so
13:43 egonw right
13:43 egonw alternatives: the CDK manager
13:43 egonw and PubChem
13:43 jonalv the Bioclipse pubchem manager has a method pubchem.downlad3d
13:44 jonalv which seems promising
13:45 * jonalv just tried: ui.open(pubchem.download3d(1)) in the Bioclispe JavaScript console and that seemed promising
13:45 jonalv just need the CID
13:46 egonw which BridgeDB has
13:46 jonalv egonw: is the the same as a CAS id do you think?
13:46 egonw either from withing the PV plugin or with the BridgeDB manager
13:47 egonw PubChem CID != CAS
13:48 jonalv because PathVision seemed to know about CAS Ids :(
13:49 jonalv or PathVisio even, right? :)
13:50 egonw well, a bit, yes
13:51 egonw why?
13:51 egonw the BridgeDB manager knows the same thing...
13:51 egonw do you have a need for CAS numbers?
13:58 jonalv egonw: You lost me. IT seems what we can get from PathVisio was CAS numbers and I was wondering how difficult it would be to dowload a 3Dstructure from that. What is BridgeDB? I think I have missed that...
13:59 egonw GPML has many identifiers...
13:59 egonw not just CAS, but HMDB, PubChem, ChemSpider, ...
14:00 jonalv egonw: oh, okej.
14:00 egonw http://www.biomedcentral.com/1471-2105/11/5
14:01 jonalv egonw: so there is a BridgeDB manager in Bioclipse or di I missunderatnd you?
14:02 egonw there is a plugin for that, yes
14:02 egonw http://jenkins.bigcat.unimaa​s.nl/job/bioclipse.bridgedb/
14:02 hybridivy egonw and jonalv , we can get the CID from  bridgedb
14:02 hybridivy http://www.helixsoft.nl/blog/?p=276
14:03 hybridivy while we are getting the SMILES i hope we can get the CID too
14:03 egonw yes, no problem
14:04 hybridivy so as jonalv suggested we could use bioclipse pubchem manger and do 3d vis part also
14:04 hybridivy :)
14:05 egonw right
14:05 jonalv that is pretty neat actually :)
14:24 hybridivy jonalv and egonw: i have been making uptodate changes to google doc which i shared, if any content is unclear or misunderstood by me , do comment on it.
14:25 jonalv hybridivy: I will try to look at it. Probably tomorrow.
14:26 hybridivy egonw and jonalv: sure do it when ever , you people find time
14:26 hybridivy :)
14:30 hybridivy joined #bioclipse
14:38 jonalv I have to run. later o/
14:53 mgerlich joined #bioclipse
14:53 cruttkie joined #bioclipse
15:18 hybridivy joined #bioclipse
15:23 hybridivy joined #bioclipse
15:58 zaetnick joined #bioclipse
17:32 looper joined #bioclipse
17:46 egonw joined #bioclipse
19:38 zaetnick joined #bioclipse
20:03 zaetnick joined #bioclipse
20:43 zaetnick joined #bioclipse

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