Perl 6 - the future is here, just unevenly distributed

IRC log for #cdk, 2014-01-30

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

All times shown according to UTC.

Time Nick Message
23:43 slyrus joined #cdk
23:52 slyrus_ joined #cdk
03:27 slyrus joined #cdk
04:51 egonw joined #cdk
05:01 slyrus hey egonw
05:05 egonw moin
05:05 zarah ni hao egonw
05:05 egonw [order] I think IUPAC is not always clear on that either...
05:06 egonw there is some guidelines on which rules are more important, but not always
05:06 egonw resulting in multiple valid names
05:06 egonw I would suggest a tree, and make it up along you go...
05:06 egonw that's what I did (with the 5 rules I implemented)
05:07 egonw in the above example, ethyl benzene may be the preferred one, with benzene being larger?
05:08 slyrus so you're just matching in order that they appear in the rules vector?
05:10 slyrus and are you basing your approach off of the html-ized "blue book" (?) or the draft Preferred IUPAC names?
05:10 egonw I have copies of print versions
05:10 egonw not of all books, though
05:10 egonw and some in Dutch
05:11 slyrus would the Preferred IUPAC names draft be a good-enough place to start/finish?
05:25 egonw since reordering of rules, if as a tree (or "priority"), should not be technically hard, but scientifically hard (manual work, expert opinions, etc), I think any start is a good start
05:26 egonw you could also consider doing an exhaustive rule matching, and allow returning both
05:26 egonw I guess it is much clearer for names that really are invalid...
05:27 egonw but your two examples are probably both acceptible..
05:29 slyrus hrm... Ok, I'm going to go by the "P-44 Seniority order of parent structures" section for now
05:34 egonw that sounds good
05:34 slyrus the next thing is getting, e.g., 2-methylbutane to work
05:38 egonw ok, heading to work now
05:38 egonw ttl
06:15 egonw joined #cdk
06:20 sneumann joined #cdk
06:21 slyrus joined #cdk
06:47 konditorn joined #cdk
07:24 egonw joined #cdk
07:39 egonw joined #cdk
08:55 jonalv joined #cdk
09:18 slyrus joined #cdk
09:30 Gpox joined #cdk
09:56 jbrefort joined #cdk
16:26 jonalv joined #cdk
16:47 slyrus egonw: around?
16:59 egonw physically, yes, mentally... less so
17:04 slyrus ok. I think we need a more involved, iterative approach to the matching, otherwise it will be rather inefficient
17:04 slyrus especially if we start thinking about retained names
17:05 egonw yes, I would expect John to have good ideas on this...
17:07 slyrus Ok, I'll start working on some emails later in the day. I think we need to both 1) decouple the finding/matching of moieties (for lack of a better term) and the prioritization of them and 2) have some sort of iterative graph-based matching approach so that we can match small numbers of nodes to build up larger named pieces
17:08 slyrus e.g. adenine and ribose => adenosine (assuming we have any hope of supporting that sort of thing)
17:09 egonw sounds good...
17:09 slyrus trying to match all possible structures to any new target structure would not work well, imo
17:09 egonw let's see how much interest we can trigger...
17:09 slyrus but coalescing various chains/rings/ring systems into nodes that can be matched might work
17:09 egonw it should get, I expect, because this is one of the missing open source cheminformatics pieces
17:10 egonw John will have excellent suggestions on how to do the matching...
17:10 egonw I love to take care of some Dutch localization
17:10 egonw and general social community building...
17:10 slyrus i'm probably going to alienate many potential contributors by doing an initial version in Common Lisp, but there's no hope of me getting this done in a finite amount of time in any other language, especially not java...
17:10 egonw getting Daniel of OPSIN involved would be good too, if only for expertise...
17:11 egonw slyrus: please talk to John
17:11 slyrus will do
17:11 egonw his graph matching work is really efficient
17:13 slyrus thanks, ttyl
18:00 sneumann joined #cdk
18:02 egonw joined #cdk
18:59 slyrus joined #cdk
19:01 jonalv joined #cdk
19:09 jonalv joined #cdk
19:11 slyrus_ joined #cdk
19:14 jbrefort joined #cdk
19:28 egonw joined #cdk
20:07 konditorn joined #cdk

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