Camelia, the Perl 6 bug

IRC log for #cdk, 2011-09-08

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

All times shown according to UTC.

Time Nick Message
03:44 sneumann joined #cdk
03:59 egonw joined #cdk
05:18 egonw joined #cdk
05:23 sneumann joined #cdk
06:33 egonw joined #cdk
07:02 Gpoks joined #cdk
07:02 Gpox joined #cdk
07:13 Gpox joined #cdk
07:46 jonalv joined #cdk
09:12 Conrad joined #cdk
10:48 sneumann egonw: have you heared news of http://www.slideshare.net/peyr​on/julio-peironcely-iccs-2011 ?
10:50 maclean joined #cdk
10:51 maclean morning (just)
10:55 egonw moin moin
10:55 egonw sneumann: yes, but only seen those slides... I have not seen the code in action, or the code itself
10:55 egonw hi maclean
10:55 maclean hi egon
10:56 egonw sneumann: so, *no*, I have not heard any *news*
10:59 * maclean would also be interested in seeing Julio's code :)
11:00 * egonw is not looking forward to another big code blob dump
11:00 maclean :/ agreed.
11:11 jbrefort joined #cdk
11:34 jonalv joined #cdk
11:42 jonalv joined #cdk
11:50 egonw maclean: I'm checking the Se patch
12:00 maclean joined #cdk
12:00 maclean egonw : takk. (was at lunch).
12:05 egonw looks good
12:06 maclean good good.
12:06 maclean I should look at S, N, and P also.
12:06 egonw thanx for updating
12:06 maclean n/p
12:07 CIA-38 cdk: maclean cdk-1.4.x * r02e65a4 / src/main/org/openscience/cdk/​dict/data/cdk-atom-types.owl :
12:07 CIA-38 cdk: Se atom type added
12:07 CIA-38 cdk: Signed-off-by: Egon Willighagen <egonw@users.sourceforge.net> - https://github.com/cdk/cdk/commit/02e​65a472f8073835cff41604f787dadfea0258f
12:07 CIA-38 cdk: maclean cdk-1.4.x * rad7753a / src/main/org/openscience/cdk/at​omtype/CDKAtomTypeMatcher.java :
12:07 CIA-38 cdk: Se atom type cases added
12:07 CIA-38 cdk: Signed-off-by: Egon Willighagen <egonw@users.sourceforge.net> - https://github.com/cdk/cdk/commit/ad7​753aa49cf01dc884ad98d3dd09e03d138fcca
12:07 CIA-38 cdk: maclean cdk-1.4.x * r55c6a83 / src/test/org/openscience/cdk/atom​type/CDKAtomTypeMatcherTest.java :
12:07 CIA-38 cdk: Se atom type test cases added
12:07 CIA-38 cdk: Signed-off-by: Egon Willighagen <egonw@users.sourceforge.net> - https://github.com/cdk/cdk/commit/55c​6a835e879b0e20b3d484b1423e99c6b91dfaa
12:07 maclean (or, rather : that's what I'm paid to do :)
12:11 maclean gah! 404 not found on a page from a workshop in March 2011 - that's some quick degeneration!
12:25 maclean egonw: do you know Martin Gütlein or the ChesS-Map software? Seems to be OpenTox related.
12:25 egonw I think he is indeed involved in OT
12:25 egonw dunno about ChesS
12:25 egonw or ChesS-Map
12:25 maclean It's quite nice.
12:29 egonw oh crap
12:29 maclean ?
12:38 egonw messed up my eclipse install by accident :/
12:38 maclean oh :(
13:04 * maclean doesn't think that a bug in some javascript is the same as a JCP bug...
13:09 egonw joined #cdk
13:22 jbrefort joined #cdk
14:44 egonw joined #cdk
14:54 egonw_ joined #cdk
15:05 maclean joined #cdk
15:10 egonw_ hi maclean
15:10 maclean hi egonw
15:10 egonw_ saw the Al patch report
15:10 maclean Ok.
15:11 maclean Am also working on : Br, Co, N, S, Ru, V, Zn.
15:11 egonw great
15:11 maclean I think Co may have been missed, but the others have various problems.
15:12 egonw it's in my rejected list
15:13 maclean oic.
15:14 egonw I'll check if I left comments
15:14 maclean Could you remind me why? I can't find a reference to it in the patch tracker comments.
15:14 maclean Ah :)
15:15 egonw cannot find any comment in the patch tracker either...
15:15 egonw weird
15:15 egonw bad me
15:15 maclean inevitable with so many patches.
15:20 * maclean starts to get the hang of git am -3
15:24 maclean Ugh. Chebi says V(OOO=O), PubChem says V(=O=O=O)...
15:26 egonw_ joined #cdk
15:32 maclean Wow, a google search for InChI=1S/4O.V/q;3*-1; hits the page for Ferrate, InChI=1S/Fe.4O/q;;;2*-1
15:33 egonw_ ugh
15:33 egonw_ google--
15:33 egonw_ they still don't get InChIs... after years of lobying...
15:34 maclean well, er, it's kind of like a substructure match though, isn't it? :)
15:34 egonw_ hahahaha
15:34 egonw_ like those guys doing similarity matching based on SMILES character strings :)
15:35 maclean Heh. Well, it would kind of work. For some things...
15:36 maclean Right, I'm deciding that [VO4]3- is indeed sp3, and has one double bond.
15:36 egonw_ no, no
15:36 egonw_ sp3 != tetrahedral
15:36 egonw_ sp3 happens to be tetrahedral
15:37 egonw_ but sp3 is a hybridization of 4 orbitals
15:37 egonw_ and thus can host 8 electrons
15:37 egonw_ 4 neighbors where one is double bonded is 10 electrons
15:37 jbrefort for V, some kind of dsp3
15:37 egonw_ Pauling turns in his grave
15:37 egonw_ jbrefort: right!
15:37 egonw_ or d2sp2
15:38 * jbrefort doesn't like hybridization
15:38 egonw_ unfortunately, some cheminfo algos need it
15:38 egonw_ I rather have it not defined, than wrongly defined :)
15:38 maclean Oh, you and your 'chemistry'. Ok, I will change it.
15:38 egonw_ maclean: which I can expect missing hybrid for weird metals
15:38 egonw_ :)
15:39 maclean d^0?
15:39 maclean (where ^ is superscript)
15:39 maclean No, wait. That's not in the atom-types.owl. Hmmm.
15:39 egonw_ that's only one d orbital, so 2 electrons
15:41 maclean there's a tetrahedral Hybridization state, shall I use that? (rather than adding "Hybridization rdf:ID="some_weird_metal"/>
15:41 egonw_ are you sure it's tetrahedral?
15:42 egonw_ 4 neighbors does not imply that
15:42 * jbrefort is
15:42 egonw_ jbrefort: got a crystal structure, by any chance?
15:42 maclean http://en.wikipedia.org/wiki/Vanadate
15:42 jbrefort isoelectronic to phosphate, btw
15:43 egonw_ so, we're looking for a hybridization with 4 orbitals that gives a tetrahedral environment...
15:43 egonw_ jbrefort: I have been wondering what the pi orbital actually looks like
15:43 egonw_ just like a pz with sp2?
15:43 egonw_ (mege of two pz's, that is)
15:44 jbrefort I don't think, probably more complicated
15:44 egonw_ yeah, interesting, isn't it :)
15:44 jbrefort note that the "pi" orbtal is not localized
15:44 egonw_ make sense, I guess
15:44 egonw_ like with carboxylic acid
15:44 egonw_ (and phosphate :)
15:45 jbrefort given the symmetry, you can use the s orbital, the three p and at least 2d
15:45 egonw_ I wish I had payed more attention in QM classes...
15:45 jbrefort so a minimum of six, one of which must be empty
15:45 egonw_ but we waited energy in reimplementing HF :(
15:45 egonw_ in Fortran
15:45 jbrefort (vanadate is not in triplet state as far as I know)
15:46 egonw_ why six?
15:46 egonw_ why triplet state?
15:46 jbrefort because in tetrahedra symmetry, d orbitals become two e orbitals and three t2
15:46 egonw_ I don't see any single electrons...
15:46 egonw_ oi, ic...
15:46 egonw_ that sounds familiar... e and t2
15:47 egonw_ jbrefort: that first free bond then... bonding, anti-bonding?
15:47 egonw_ or, umm... neither...
15:47 jbrefort I don't know
15:47 egonw_ non-bonding...
15:48 egonw_ wondering if vandate would be back-donation :)
15:48 jbrefort most probably yes
15:48 jbrefort if this has any meaning
15:49 egonw_ well, it's already quite charged, so likely not too keen on even more electrons...
15:49 egonw_ maclean: still with us?
15:49 * egonw_ can just barely follow...
15:49 maclean Heh. Oh, yes : yes. Back donations! Oh my yes.
15:49 maclean (no : not at all... :( )
15:50 * maclean is looking at http://www.crystallography.net
15:50 egonw__ joined #cdk
15:50 egonw__ grpmh... ???
15:50 egonw__ I mean, the bus is bloody stuck in a traffic jam...
15:50 egonw__ how can it loose wifi???
15:52 maclean Cars are scattering the signal due to the multiple overlapping Faraday cage effect. :P
15:52 maclean Also : http://www.ebi.ac.uk/thornton-srv/databases/cgi-bi​n/pdbsum/GetPage.pl?pdbcode=n/a&amp;template=het2p​db.html&amp;param1=VO4&amp;s=27825884&amp;o=OFFSET
15:52 zarah maclean's link is also http://tinyurl.com/3hs8pec
15:53 maclean Actually, I've seen tetrahedral vanadate tons of times in proteins, as it looks like phosphate (you mentioned this already).
15:53 egonw yeah, well, I'm convinced... please do add such references to the unit test...
15:54 egonw really, that helps a lot
15:54 egonw but, ummm...
15:54 egonw so, we can finally get rid of this #arseniclife stuff, and move on to #vanadiclife ?
15:55 maclean Heh. Arsenic based life is so 6 months ago
15:55 egonw :)
15:55 jbrefort http://www.crystallography.net/cif/1/1001155.cif shows perfectly tetrahedral vanadate
15:55 egonw ah, that's OCD, right?
15:56 egonw COD
15:56 egonw yeah, I'm so happy they are teaming up with CrystalEye...
15:56 egonw first, huge free chemistry database!
15:56 egonw so, my summer project next year, will be to mine these CIFs for CDK atom types
15:57 egonw (non trivial: missing bond orders, probably missing hydrogens :)
16:00 jbrefort just missing bonds ;)
16:01 egonw CDK has a RebondTool
16:02 maclean Missing bond orders eh... Time for my line graph coloring algorithm! :)
16:02 egonw which does a pretty good job
16:02 egonw would be a nice data set to machine learn atom types
16:03 egonw with a proper, absolute yet alignment-free stereochemistry around atoms
16:04 maclean What do you mean by 'alignment-free'?
16:05 egonw with any atom, you don't know where is north, up, south, or leftish
16:05 egonw quite like the CIP algo problem
16:05 maclean Ah.
16:06 maclean I will be interested at some point in enumerating stereochemistry. Not sure if that's relevant, just saying is all.
16:09 maclean Right. I'm off for some vending machine tea (pluses : 20p, minuses : slightly nasty).
16:10 egonw :)
16:20 egonw crap... 100 minutes on the bus now... 17km away from where I started
16:28 jbrefort not very quick indeed
16:30 egonw_ joined #cdk
16:34 egonw__ joined #cdk
16:35 maclean oh btw egon, do you still want to be cc'ed on mails between Aaron and I on molecule enumeration?
16:39 maclean oh, g2g.
16:47 egonw__ joined #cdk
16:49 egonw joined #cdk
16:52 egonw_ joined #cdk
17:05 egonw__ joined #cdk
18:22 egonw__ joined #cdk
20:04 CIA-38 cdk: mark_rynbeek * r15623 /jchempaint/branches/3_2/build.xml: added reacion jar

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