2016.08.19 Community Chat

From Eterna Wiki

Timestamps are UTC -5

[13:56:54] <ChatBot> KA-BOOOM!!!
[14:00:24] <jandersonlee> launch - I mean lunch.
[14:01:31] <jandersonlee> Is this the palyer chat week or the dev chat week?
[14:02:33] <hoglahoo> not sure :/
[14:03:18] <hoglahoo> It seems like I heard something about it not being as regular lately
[14:04:41] <Nando> hi guys
[14:04:52] <hoglahoo> hey nando
[14:04:58] <whbob> Hi.
[14:05:06] <Poll-na-gColm> howdy
[14:05:42] <MasterStormer> hi
[14:06:11] <jandersonlee> Hi Nando, Hoglahoo, whbob, et al
[14:08:49] <whbob> rhiju was here last week.
[14:09:06] <whbob> no chats saved for a few weeks.
[14:09:15] <jandersonlee> probably a player week then
[14:09:47] <jandersonlee> Nando - any news on the last lab results? I hear the data is coming in
[14:10:55] <Nando> that's what I'm hearing as well, sorry to disappoint but I don't know more than that
[14:11:03] <jandersonlee> np
[14:13:56] <whbob> In the present lab, the kernal attractors that switch a lot are like goldie lock's. Can't be too strong and can't be too weak. Just right is waht they want :)
[14:14:11] <whbob> *what
[14:14:39] <whbob> Not a lot of variations available :(
[14:18:43] <jandersonlee> It seems tougher than the first round.
[14:21:06] <whbob> We haven't seen round 2 dat for the A/B labs yet. Then comes the two TB rounds before we see how those oligo's like our designs.
[14:22:01] <whbob> 85 bases tight as a violin string.
[14:22:43] <Poll-na-gColm> Will there be another posted after round 2 finishes?
[14:23:26] <jandersonlee> No doubt, but it may be a week or two. Nondo, any idea how soon after the lab closes?
[14:23:39] <jandersonlee> s/Nondo/Nando/
[14:24:53] <Nando> so far, I haven't heard anything about the next steps... I assume we're waiting for experimental results to make decisions
[14:25:13] <Nando> or the end of summer and the returns from vacations :D
[14:25:26] <whbob> :)
[14:26:41] <jandersonlee> Might it be new oglios or back to the first set? If the openTB lab1 resuts are nearly in I'd prefer to reuse those oglios for the next lab.
[14:27:13] <whbob> I've forgotten if the oligo's are from RNA or DNA, but if one type doesn't work they still have the others to try.
[14:27:52] <jandersonlee> I believe the oglios are tRNA
[14:27:59] <Nando> we're working with RNA exclusively at the moment, DNA would be a possibility, if I recall correctly what Johan said
[14:28:44] <Jennifer-Pearl> hi
[14:28:56] <whbob> Hi Jen
[14:29:03] <jandersonlee> hi JP
[14:29:05] <Jennifer-Pearl> sorry I am late
[14:29:23] <jandersonlee> better late than never
[14:29:39] <Jennifer-Pearl> i am trying to show up more often again.
[14:30:10] <whbob> @ jen: did you put any of your code on github?
[14:30:59] <Jennifer-Pearl> no. I have been very busy with life
[14:31:29] <whbob> that's what happens when you are making plans :)
[14:31:40] <Jennifer-Pearl> i have only posted the VS solution on Get Sat
[14:32:46] <Jennifer-Pearl> i have an idea for a new way to visualize the data to get away from pairing probabilites when looking at point distribution for specific stacks
[14:33:09] <whbob> I was looking at eterna's stuff on github and didn't see anything on dotp;ot stuff so I wondered.
[14:33:10] <Jennifer-Pearl> better phrasing would be a better way to visualize it
[14:34:07] <Jennifer-Pearl> i need to do it
[14:34:28] <Jennifer-Pearl> working on some anaysis at the moment
[14:34:42] <whbob> So, you have a new way to graph?
[14:35:15] <Jennifer-Pearl> yeah i think i want to grah by number of designs with stack
[14:35:48] <Jennifer-Pearl> with a specific stack vrs by pairing probability when looking at eterna scores
[14:36:15] <Jennifer-Pearl> that way when you graph it you can see teh distrobutions naturally
[14:37:47] <whbob> I know you were doing earlier puzzles, but can you think of oligo's and their kernal attractors a stacks ?
[14:40:23] <whbob> that would be interesting
[14:48:39] <whbob> I've been trying to make a fusion table out of the R103 spreadsheet, but I don't know how to use the PI to get the projecy and puzzle ID's with the design ID's so I can merge two sheets
[14:48:53] <whbob> * API
[14:49:05] <whbob> *project
[14:51:00] <jandersonlee> you can merge two fusions tables, or export to .csv do it in Excel/LbreOfficeCalc/etc and make a new fusion table
[14:52:55] <whbob> do I have to manually put the project and puzzle ID's into a spreadsheet with the design ID"'s?
[14:54:02] <Jennifer-Pearl> sorry was trying to order food and my chat crashed
[14:54:14] <jandersonlee> or write/modify an EternaScript to do so
[14:55:06] <Jennifer-Pearl> im a bit behind on terminology but what is the kernal attractor. Is that the nucs that the oligo binds to?
[14:55:28] <jandersonlee> http://www.eternagame.org/web/script/6046883/ could be a starting point
[14:55:58] <jandersonlee> @JP, yes I belive that is the meaning
[14:56:40] <Jennifer-Pearl> in that case yes you can think of th eoligos and the kernal attractors as stacks
[14:58:25] <Jennifer-Pearl> with the current release of it you can look at single oligo labs and if you seach for a couple known binding site nuc pairs you shoul dbe able to look at the various oligo/kernal attractor combinations
[14:58:53] <Jennifer-Pearl> that are possible and then analyize them the same way I look at mandatory binding sites
[14:59:16] <Jennifer-Pearl> should be able to predict what configurations are the best historicaly
[14:59:26] <Jennifer-Pearl> or more look at
[15:00:40] <Jennifer-Pearl> but I bet any predictions you make off of that would be pretty acurate if the plots look like the ones for the special stacks I showed at Eternacon
[15:01:45] <Jennifer-Pearl> NUPACK development is going slow right now though so cant do more than one oligo for now
[15:04:07] <whbob> @ jal: just saved that script to look at it later. Thanks
[15:05:55] <whbob> It's amazing how whishful developments are cropping up fast while the project puzzles are solwoin to a snails pace.
[15:06:09] <whbob> *slowing
[15:07:16] <jandersonlee> what is slowing? lab rounds or lab tools?
[15:07:23] <whbob> rhiju was saying last week that the future could require much more computational power & development cycles.
[15:08:05] <whbob> the speed of seeing the result a a single base mutation
[15:08:34] <jandersonlee> yes. running NuPACK on a 6 strand design can take a minute on a desktop machine
[15:09:03] <whbob> not to mention the time to get johan's spread sheets
[15:09:06] <jandersonlee> better than my first attempt, which took 3 hours
[15:10:05] <jandersonlee> lunch is over - later folks.
[15:10:22] <whbob> I'm not complaining, quality takes time :) I just have a hard time concentrating while a mustation completes its computation.
[15:10:42] <whbob> *mutation
[15:10:50] <whbob> BY jl
[15:14:08] <Jennifer-Pearl> bye JL
[15:15:32] <whbob> there was a funny story about an owner whose dog lost concentration when it saw a squirrel. when I make a mutation, I see squirrels :)