2013.12.18 Dev Chat

From EteRNA WiKi
Jump to: navigation, search

( Link back to Dev Chat archive )

 

On Labs & Barcodes

 

jeehyung: hello all [3:00 PM]

Eli Fisker: hi Jee :) [3:00 PM]

lroppy: Hi jee [3:00 PM]

jeehyung: welcome to Eterna bi-weekly dev chat : ] [3:00 PM]

Eli Fisker: thx [3:00 PM]

hoglahoo: yay, thanks. hello [2:00 PM]

machinelves: hey! :) [3:00 PM]

jeehyung: Hi Eli, lroppy, hoglahoo, LFP [3:01 PM]

Eli Fisker: hi Machine [3:01 PM]

machinelves: hi! :) [3:01 PM]

Eli Fisker: :) [3:01 PM]

jeehyung: hi machielves [3:02 PM]

jeehyung: Everyone as always please throw in suggestions & questions : ] [3:02 PM]

machinelves: hey jee! my main followup this week is the schedule for chat guidelines to be posted, and many thanks to the team for all their hard work! [3:02 PM]

JR: hi -what do you think is max lab length? [3:03 PM]

jeehyung: @machinelves the chat guideline is being discussed - we'll probably come up with something next week. Thanks for all the suggestions. [3:04 PM]

machinelves: great, thank you! :D [3:04 PM]

jeehyung: @JR we still don't know yet. We tried another method to synthesize RNAs up to 150 bases. Hopefully it'll work this time [3:04 PM]

hoglahoo: Is there another google live meeting being planned? [2:04 PM]

JR: still trying -great [3:04 PM]

jeehyung: @hoglahoo yes - it'll be once a month.  The first friday of every month. [3:05 PM]

hoglahoo: neat [2:05 PM]

Eli Fisker: :) [3:05 PM]

machinelves: @jee another low priority followup is: should I continue working on an Issue Tracking Roadmap like this: http://eternawiki.org/wiki/index.php5/Issue_Roadmap  [3:05 PM]

machinelves: or is your team intending to use another system for tracking priority of the bug queue? [3:06 PM]

JR: any upgrades in the works? [3:06 PM]

JR: other than bugs [3:06 PM]

machinelves: since the spreadsheet is deprecated, and github has a queue but no apparent prioritization ( correct me if wrong ) [3:07 PM]

machinelves: i.e. looking for central roadmap for EteRNA priority & status of issues [3:07 PM]

lroppy: ditto [3:07 PM]

Eli Fisker: Yes, it will be nice to be sure that we see a fix, when it is made [3:09 PM]

Brourd: hey Jee [3:09 PM]

lroppy: e.g. just found a bug in puzzlemaker [3:09 PM]

rhiju: hi all (and hi jee) [3:09 PM]

Brourd: hey Rhiju [3:09 PM]

Eli Fisker: hi Rhiju :) [3:09 PM]

lroppy: Hi Rhiju [3:09 PM]

machinelves: Eli makes a good point, it is not just to fix priority, but also to ensure status is centrally reported so that players know when a github ticket is closed out [3:09 PM]

jeehyung: @machine thanks for bringing that up again. We'll definitely re-visit this once we are ready for open dev [3:10 PM]

jeehyung: hi Rhiju [3:10 PM]

rhiju: not sure where jee went. i wanted to login earlier but chat is not letting my laptop join in. had to move to another terminal [3:10 PM]

JR: bugs are no big deal - just write report them in forum [3:10 PM]

rhiju: (oh hi jee) [3:10 PM]

machinelves: @jee understood, thank you!  [3:10 PM]

jnicol: hi everyone [3:10 PM]

jeehyung: @rhiju I believe you had topics in mind - do you want to proceed? [3:11 PM]

Eli Fisker: hi John :) [3:11 PM]

rhiju: yea, as usual, i wanted to brainstorm about how to get papers written [3:11 PM]

Eli Fisker: Ok [3:11 PM]

rhiju: i think the eterna live group meeting is a good step [3:11 PM]

rhiju: especially if we can get more player participation [3:11 PM]

machinelves: @jee, fyi issue roadmap began before concept of open dev, to streamline dev-side understanding of priority / status, not just player-dev potential. but i do see that player-dev issue reporting will require same tools [3:12 PM]

rhiju: and perhaps if we can inspire player teams to eventually set up their own hangouts [3:12 PM]

JR:  could make a list of where we could focus [3:12 PM]

machinelves: hi rhiju! [3:12 PM]

machinelves: thanks again for the stanford chat, that was super cool :D [3:12 PM]

rhiju: @machinelves, thanks for taking part and for asking questions [3:12 PM]

jandersonlee: hypothesis->expDesign->expt->analysis->paper [3:12 PM]

rhiju: @jandersonlee, yup [3:12 PM]

Brourd: Sounds so easy. [3:13 PM]

machinelves: happy to help :) [3:13 PM]

jandersonlee: need hypothesis and a way to test them [3:12 PM]

Eli Fisker: hi JL :) [3:13 PM]

lroppy: could make labs more structured [3:13 PM]

rhiju: yes, i think the current lab setup is doing a reasonable job of keepting track of hypotheses and results [3:13 PM]

rhiju: but could be improved [3:13 PM]

rhiju: so let;s discuss that [3:13 PM]

lroppy: so they are more natural stepping stone to papers [3:13 PM]

rhiju: @lroppy, yea! [3:13 PM]

machinelves: one note on lab submission: if there are constraints that must be followed lest the results become unusable, it would help to itemize these instructions in a clear linear fashion [3:14 PM]

lroppy: fewer labs would help [3:14 PM]

lroppy: actually [3:14 PM]

machinelves: ah, exactly lroppy [3:14 PM]

rhiju: ok, that's very interesting [3:14 PM]

lroppy: more time to work on fewer [3:14 PM]

machinelves: nando luckily found this issue in a recent lab, and it was redone [3:14 PM]

rhiju: @machinelves, ok that needs to be fixed. [3:14 PM]

jandersonlee: how about a braintom area on the wikie wher people can flesh out hypotheses before making a lab proposal [3:14 PM]

rhiju: jee just worked out a way to embed 'notes' in the puzzle [3:14 PM]

jeehyung: I'll note these suggestions  [3:15 PM]

machinelves: like very very clear: 1. Must place stronger bonds in first NNNN loop to get points. 2. The purpose is to test which loop bond nature prefers... etc.... [3:15 PM]

Brourd: Great [3:15 PM]

Eli Fisker: Cool, Rhiju [3:15 PM]

lroppy: lab could also include a sign-up process for those will to help [3:15 PM]

rhiju: @jandersonlee, well how about just only allowing fewer labs per round -- say 8 [3:15 PM]

lroppy: analyze [3:15 PM]

machinelves: because players are used to quickly running through puzzles at lightning speed, and may not stop to read whole instruction block [3:15 PM]

rhiju: but each one gets more slots [3:15 PM]

lroppy: and author results [3:15 PM]

Eli Fisker: And cool Jee, on making embedde d notes possible [3:15 PM]

lroppy: think of lab more like mini-reseearch project [3:16 PM]

rhiju: back to papers [3:16 PM]

lroppy: with its own team [3:16 PM]

rhiju: @lroppy yes [3:16 PM]

machinelves: also, at this point does the lab have capacity to run all designs? or is voting still necessary to choose? [3:16 PM]

Eli Fisker: Rhiju, I will definetely like the more slots on each lab [3:16 PM]

jandersonlee: quality over quantity yes, but how best to improve the quality? [3:16 PM]

jandersonlee: @Eli, just ask for more [3:16 PM]

rhiju: @jandersonlee, how about really making this a competition [3:16 PM]

Eli Fisker: Some labs only got 20-25 and often it would have been interesting to see more [3:16 PM]

hoglahoo: winner gets a das lab tshirt [2:16 PM]

Eli Fisker: JL, I do [3:17 PM]

rhiju: like each team that gets synthesis submits a paper [3:17 PM]

rhiju: i could help define a template for such papers [3:17 PM]

lroppy: use incentives [3:17 PM]

lroppy: I hate to say it [3:17 PM]

machinelves: w00t! templates!!! [3:17 PM]

Brourd: Eh. @ Eli - are you referring to my own projects? Because those are only 20-25 slots for a reason :) [3:17 PM]

lroppy: but points could tie in [3:17 PM]

rhiju: and we could have a 'panel' decide who 'wins' [3:17 PM]

rhiju: and the prize could be t-shirts [3:17 PM]

Eli Fisker: Brourd, I'm saying more results will be interesting on those [3:17 PM]

rhiju: but also serious help in getting the papers published [3:17 PM]

rhiju: in fact even if the player papers don't 'win' they could be deposited and indexed [3:18 PM]

JR: if you attach notes to lab - and have devs vote on if interesting direction or not [3:18 PM]

rhiju: for example i'm very excited about this: http://biorxiv.org [3:18 PM]

Eli Fisker: I will be ineterested in getting help on writing a paper [3:18 PM]

Brourd: More results is just what it is. They are called pilot runs for that very reason. [3:18 PM]

rhiju: @JR, devs will have trouble keeping up with even 8 labs. [3:18 PM]

rhiju: (or even 1 lab!) [3:18 PM]

jandersonlee: I think it might help to have "paper labs" where teh hypotheis and experimental design are laid out and vetted first (but maybe not all labs that way) [3:18 PM]

rhiju: ok, how about some set of pilots that get ~20 synthesis slots [3:19 PM]

Eli Fisker: The labs with very few slots, causes there to be many more labs for each round [3:19 PM]

lroppy: agree with jal [3:19 PM]

rhiju: and then paper labs which get ~200 designs/cycle [3:19 PM]

jandersonlee: @rhiju sound fair [3:19 PM]

rhiju: @jal, yea i agree. calling them 'paper labs' is a good idea too. thanks [3:19 PM]

jandersonlee: but it would haelp to have coaching in teh design of the paper labs [3:19 PM]

Brourd: Oh yes, Rhiju. I had a question. Is there any possibilty that EteRNA players can be given access to the major journals and such? [3:20 PM]

lroppy: start with template [3:20 PM]

machinelves: i second brourd's request [3:20 PM]

rhiju: @brourd, might be tough to get papers [3:20 PM]

lroppy: need an interpreter for us [3:20 PM]

machinelves: like portal to journals within eterna login? [3:20 PM]

JR: how will players know if barking up old or wrong tree   [3:20 PM]

rhiju: so instead we should allow paper labs to have several team members [3:20 PM]

rhiju: and each team could recruit someone with access to papers [3:20 PM]

machinelves: if this is hosted by university, does their access extend to us? [3:20 PM]

rhiju: ;) [3:20 PM]

Brourd: Can't use Stanford credentials to give players access? [3:21 PM]

rhiju: @machinelves, exactly. [3:21 PM]

rhiju: @brourd, i'm guessing no. [3:21 PM]

rhiju: the other thing -- coaching .... [3:21 PM]

Eli Fisker: Libraries sometimes gives acces to more papers [3:21 PM]

rhiju: for this to scale, we need to recruit folks to each team to help in writing. [3:21 PM]

lroppy: i think many of us actually have access [3:21 PM]

lroppy: via alumin association of colleges/universities [3:21 PM]

rhiju: i will likely be teaching or co-teaching some classes next year where stanford students will need practive in writing and might team up with eterna plauers [3:21 PM]

lroppy: *alumin [3:22 PM]

rhiju: but that won't work year-round [3:22 PM]

rhiju: @lroppy, yes [3:22 PM]

Eli Fisker: Rhiju, that sounds interesting [3:22 PM]

lroppy: y [3:22 PM]

machinelves: re: incoming NOVA, i understand there will be separate server. two points: 1) no chat there i understand, but may lead teachers/parents to think main site has no chat. thoughts? [3:23 PM]

rhiju: ok. ill think about a template and'rules' over the next week or two, and discuss with devs. [3:23 PM]

jandersonlee: well some help in lab design (if not a coach) [3:23 PM]

machinelves: 2) could issue request for teacher volunteers to moderate chat, to take pressure off players to manage trolls etc... [3:23 PM]

jandersonlee: @machinelves - yes - encourage teachers to be online during their classes [3:23 PM]

lroppy: they are usually on when their classes are first introduced to it [3:24 PM]

lroppy: don't think this is big issue [3:24 PM]

jeehyung: @machine will take a note. thanks [3:25 PM]

machinelves: kk thanks! [3:24 PM]

rhiju: folks, i gotta go in 5-10 mins -- any questions regarding experiments or papers? [3:26 PM]

machinelves: @jee followup on issue I think you resolved, but to confirm: for player dev, all personal PM data should be sanitized from branch [3:26 PM]

Brourd: How is the Das lab dealing with the RNA length problem? Is there a fairly positive chance of success? [3:27 PM]

jandersonlee: suggest max 1-3 paper labs per round [3:27 PM]

rhiju: @brourd, we have reordered the templates from a different company, as well as from our faovrite source (but without some "padding" with strings of A's). [3:28 PM]

rhiju: we will try both [3:28 PM]

jandersonlee: and teams/groups would be great as lab admins [3:28 PM]

Brourd: ah [3:28 PM]

Eli Fisker: Good luck with it, Rhiju [3:28 PM]

rhiju: @brourd, if it doesnt work, ill go back onto the bench. [3:28 PM]

machinelves: @rhiju does the lab deal with methylated RNA vis a vis epigenetics? [3:29 PM]

rhiju: @jandersonlee, yes i was thinking the same thing. 8 is too much [3:29 PM]

lroppy: try no more than 2 [3:29 PM]

jnicol: rhiju, I am interested in a self replicating RNA lab and would love to write a paper on it :) how would this process start? [3:29 PM]

machinelves: just read there are more than 4 bases if you count variations on methylated bases [3:29 PM]

rhiju: @machine, not my lab, but its fascinating stuff. the hot topic is actually methylation of DNA, but that doesn't rule out RNA methylation (or other modifications) encoding informations that is passed from cell to progeny [3:30 PM]

Eli Fisker: Oh, cool, Machine. I heard about methylated DNA, but not methylated RNA [3:30 PM]

Eli Fisker: But of cause. 3 of the bases re the same [3:30 PM]

Eli Fisker: are [3:30 PM]

machinelves: ya i wasn't sure about the RNA vs DNA contention, i am n00b [3:30 PM]

Eli Fisker: Well, actually I think there was methylation that helped with folding the ribosomal rna [3:31 PM]

rhiju: i heard recently that male mammals encode the environmental status of food through the number of tRNA fragments in sperm.( but i haven't read those papers yet.) lots of possibilities of passing on biological information to offspring via RNA. mostly unexplored [3:31 PM]

rhiju: @machine, there are at least 100 RNA bases in human cells [3:31 PM]

rhiju: crazy!  [3:31 PM]

machinelves: 0.o [3:31 PM]

jandersonlee: afk [3:31 PM]

rhiju: @jal and @lroppy, very good. Two. [3:32 PM]

machinelves: intriguing [3:32 PM]

rhiju: How about 'expert' projects? [3:32 PM]

Eli Fisker: Here was the article I was thinking about:  [3:32 PM]

Eli Fisker: http://phys.org/news/2013-10-choreographed-origami-ribosomal-rna-requires.html [3:32 PM]

machinelves: thanks Eli! :D [3:32 PM]

lroppy:   how is expert project different than paper project? [3:32 PM]

rhiju: at any given time, we could have several 'pilot' labs (perhaps not seen by default), 2 'paper labs' [3:33 PM]

Eli Fisker: And yes, crazy, Rhiju. :) [3:33 PM]

Eli Fisker: But fascinating that such world exist [3:33 PM]

rhiju: an expert project is led by an acaedmic (but not the das lab) [3:33 PM]

Eli Fisker: np, Machine [3:33 PM]

Brourd: So, we have two paper projects per round. Would there need to be a paper written based just on the results for one round? [3:34 PM]

lroppy: @brourd - should permit follow-ons [3:34 PM]

rhiju: @brourd, i think it would be tough to have papers written based ona single round. [3:34 PM]

rhiju: @lroppy, yea right, followons. [3:34 PM]

lroppy: ;-) [3:34 PM]

JR: where would i look if i wanted to learn more about shape data? [3:34 PM]

Brourd: Right, so, we have two paper projects oer round, but how many paper projects in total waiting? [3:35 PM]

Eli Fisker: @Rhiju, for my part I think most of my paper ideas are not based on a particular lab, but rather the lab data from earlier labs [3:35 PM]

rhiju: @eli, ok what would you propose? [3:35 PM]

machinelves: you know, at this point we are gathering so much lab data, we could have an EteRNA journal library of our own, lol [3:35 PM]

lroppy: need to formalize an open (maybe weekly) forum (like the dev chat) for folks to talk over papers and ideas [3:35 PM]

Eli Fisker: I think that lab with hypothesis and then resulting in paper is a good idea [3:35 PM]

Brourd: @ JR - plenty of papers have been written, and Rhiju himself has written a few forum posts that shed a little light on the subject. [3:36 PM]

Eli Fisker: Just that other roads might be needed [3:35 PM]

Eli Fisker: too [3:36 PM]

lroppy: agree with Eli.  cannot just use lab [3:36 PM]

rhiju: ok, last question from me: [3:36 PM]

rhiju: for those of you who are setting up labs, how important has it been to get designs from players> [3:37 PM]

Brourd: Depends on the design I guess. [3:37 PM]

rhiju: that is, could you be doing just as well if you were 'granted' 20 slots (or 200) and then coordinated with some folks to figure out what to submit -- rather than having a playable puzzle? [3:37 PM]

Eli Fisker: For my part I like that others try out solve the shape. I get more creative solves than if only I should solve it.  [3:37 PM]

Brourd: Normally, players have a pretty good idea as to what they should do. [3:38 PM]

machinelves: that is interesting... i have wondered if we need 3 submissions per player, but the whole point of crowd-sourcing is to cash in on group intelligence [3:38 PM]

rhiju: @brourd, you mean the players who set up the projects have a good idea of what they really want submitted? [3:38 PM]

Eli Fisker: It also gives more varied data for further analysis later [3:38 PM]

jandersonlee: there could be some labes where you just want to try all 36 variants of a quad [3:39 PM]

machinelves: even outliers provide boundaries. without player contribution, results will ipso facto get very tight & clustered [3:39 PM]

rhiju: @eli, i agree. but maybe we should award lots of little 'pilots' or 'startups', and then take advantage of eterna's collective intelligence mainly for 2 'paper labs' per round. [3:39 PM]

Brourd: No, I mean the "crowd" players. The project admins, as Eli said, may become a bit close minded, less likely to use other solutions, etc. [3:39 PM]

rhiju: @brourd, ok thanks [3:39 PM]

jandersonlee: but most interesting data comes from unusual player input [3:39 PM]

Brourd: Correct [3:40 PM]

Eli Fisker: Yes, I agree with JL [3:40 PM]

lroppy:  and most of those do not get synth [3:40 PM]

JR:  agree with jl [3:40 PM]

jandersonlee: or Brourd's fixed bases ;) [3:40 PM]

lroppy: hence need for more slots [3:40 PM]

lroppy: per lab [3:40 PM]

machinelves: i see what you mean about wanting to test an explicit set though [3:40 PM]

Brourd: Well, I don't think all my projects have locked residues. [3:40 PM]

rhiju: @lroppy, interesting. how many slots? [3:40 PM]

machinelves: maybe both, like you said [3:40 PM]

jandersonlee: but all slots are not getting taken at 3 sumbmissions per player in some cases [3:40 PM]

lroppy: @rhiju - depends on project [3:40 PM]

lroppy: ;-) [3:40 PM]

lroppy: but agree 20 is often too few [3:41 PM]

lroppy: if science is observation, then larger #s of observations helps [3:41 PM]

jandersonlee: <50? [3:41 PM]

rhiju: ok, what if it was easy to get, um, 20-40 slots/round, but more competitive to get 500 slots for paper labs? [3:41 PM]

jandersonlee: for pilot? [3:41 PM]

rhiju: @jal, yea 20-40 for pilot [3:42 PM]

lroppy: maybe fewer on paper labs, reserving some of those 500 slots for subsequent rounds of earlier paper labs [3:42 PM]

jandersonlee: N admin slots + K player slots (man need more than 3/player at 500) [3:42 PM]

jandersonlee: *may need* [3:42 PM]

Brourd: So, when would this "paper lab" system be launched? [3:43 PM]

rhiju: @jal. yes, good. 100 admin + 400 player total? [3:43 PM]

jnicol: i vote for the 500 slots and customized barcodes for a paer lab [3:43 PM]

Brourd: If we had to give a general time frame. [3:43 PM]

rhiju: @brourd, it could be in january. [3:43 PM]

machinelves: instead of votes choosing slots, could run all submissions against algorithm[s] first, then choose distribution representing mostly strong folders, but also some outliers? [3:43 PM]

rhiju: we really want to get some momentum towards writing publications [3:43 PM]

Brourd: I was meaning to ask about the customized barcodes. [3:43 PM]

Eli Fisker: The reason, I prefer more slots pr lab generally is, because when I want to analyse a particular element. I will rather have 100 designs with data on that particular spot, so there is a good chance it shines through what will be good to do.  [3:44 PM]

jandersonlee: how about round robbin through player submissions till slots used [3:44 PM]

rhiju: @jee, anyprospect of leaving barcodes totally open for lab creators and admins? we just need 8 nucleotides to beunique at the end. [3:44 PM]

Brourd: @ Rhiju - when it comes to barcodes that are completely unpaired, does that affect the SHAPE probing or synthesis at all, with all those snippets of barcode sequences in the solution? [3:44 PM]

jnicol: and probably run the paer lab by itself, not in the cloud with all other labs [3:44 PM]

jandersonlee: voting for 400 slots boggles the mind! [3:44 PM]

machinelves: i agree with janderson... some randomization of selection is needed for more accurate depiction of results [3:44 PM]

rhiju: @brourd, the real experts on barocde interference  are some of the player s(including some logged in!), not me [3:45 PM]

machinelves: also, players would have more time to be thoughtful on lab design if not spending so much time also voting [3:45 PM]

lroppy: long-standing request to have modified stats sans bar-code [3:45 PM]

rhiju: @machinelves, perhaps it could be up to the lab 'admins' to decide how much to take voting into account. [3:45 PM]

Brourd: Yes, but it is more of a question of the sequencing, since there would be a considerable number of partial barcode sequences that cannot be attributed to any RNA sequence. [3:46 PM]

machinelves: @rhiju interesting thought! [3:46 PM]

Eli Fisker: Is it a question about if those loose barcode snippets pair up with each other? [3:46 PM]

rhiju: hey everyone i have to go. thanks for a great chat and lots of brilliant suggestions! [3:47 PM]

machinelves: thanks rhiju! :D [3:47 PM]

Eli Fisker: Bye Rhiju [3:47 PM]

Brourd: I believe there would not be 400 slots for a single target though, but split up among five, or ten different targets. [3:47 PM]

jandersonlee: tx R [3:47 PM]

Brourd: Thanks Rhiju [3:47 PM]

lroppy: bye &Thx R [3:47 PM]

jandersonlee: gtg [3:47 PM]

Eli Fisker: bye JL [3:47 PM]

machinelves: bye! [3:47 PM]

mat747: hi [3:47 PM]

Eli Fisker: hi Mat :) [3:48 PM]

machinelves: hey mat! [3:48 PM]

lroppy: hey mat [3:48 PM]

mat747: any got a chat log [3:48 PM]

mat747: anyone [3:48 PM]

machinelves: ya [3:48 PM]

Eli Fisker: Yes. Coming your way [3:48 PM]

jeehyung: @mat I'll post a chat log soon [3:48 PM]

mat747: hi machine [3:48 PM]

jnicol: jee, any progress on the documentation for the open development source code? [3:48 PM]

mat747: eli - Jee,  thx [3:48 PM]

machinelves: hey hey :) [3:48 PM]

Eli Fisker: Mat, in your mail [3:49 PM]

machinelves: @jnicol they're still working on player dev if that is what you mean [3:49 PM]

Brourd: @ Eli - the barcode is a form of identification. What SHAPE does is modify the residues, which should stop reverse transcription. In essence, those various nucleotides that make up the barcode, when unpaired, will most likely be modified by SHAPE, and I am simply wondering if that affects any of the various steps in synthesis, or the quality of results. [3:49 PM]

mat747: eli - thx [3:50 PM]

lroppy: gtg adios all [3:50 PM]

Eli Fisker: bye Iroppy [3:50 PM]

machinelves: bye! [3:50 PM]

mat747: bye lroppy [3:50 PM]

Eli Fisker: @Brourd, good question. I am not totally sure I understand [3:51 PM]

Hyphema: Brourd. perhaps you can test that thought in my lab "break the Barcode" [3:51 PM]

Eli Fisker: Could you try explain it more [3:51 PM]

machinelves: @Eli/Brourd I too wonder about the effects of barcode nts on the rest of the molecule [3:51 PM]

Hyphema: which of course was JR's original lab [3:51 PM]

Eli Fisker: Machine, there are effects of the barcode on the rest of the design [3:51 PM]

Eli Fisker: Especially if it actually folds [3:51 PM]

Brourd: Well, when unpaired, machinelves, there is always the potential for the barcode to participate with other residues to become a part of the mfe. [3:52 PM]

machinelves: right, especially for unpaired. and even pairs have energetic effects, or do i misunderstand? [3:52 PM]

Eli Fisker: Machine, I think that a stem that is part of the design, the design is acting back on it, and the barcode is acted on by the design [3:53 PM]

Brourd: When  it is a hairpin stem, most interactions would be based on the geometry of the RNA, with tertiary interactions, or just the barcode stem causing some form of displacement. [3:54 PM]

Eli Fisker: It is not a neutral thing that an extra hairpin is there [3:53 PM]

machinelves: exactly [3:54 PM]

Eli Fisker: It aint invisible [3:54 PM]

machinelves: hmm, we could run labs with the winning designs *with barcode* against same designs *without bardcode* ? [3:54 PM]

machinelves: for a control? [3:54 PM]

Eli Fisker: Yes,  [3:55 PM]

mat747: any news on when diana is going to have the beta lab UI out for testing ? [3:55 PM]

Brourd: what is a winning design? [3:55 PM]

machinelves: highest rated? i am still pretty new to labs, so correct me as i go [3:55 PM]

Eli Fisker: Mat, unfortunately not [3:55 PM]

Brourd: There are various other factors than the barcode that can cause SHAPE signatures to differ. [3:56 PM]

machinelves: @mat tickets have individual status https://github.com/EteRNAgame/EteRNA-public/issues [3:56 PM]

Eli Fisker: Machine, good idea. I suggested that to Rhiju, for a check on how SHAPE data would differen between our past labs and our current cloud labs [3:57 PM]

machinelves: @brourd interesting! it is good to keep an eye on these, so that we can followup on the research as bandwidth opens up [3:57 PM]

Brourd: @ Eli - there were a few "expert" projects that checked that. [3:58 PM]

machinelves: @Eli interesting, thanks! also, what does cloud lab mean? [3:58 PM]

Eli Fisker: Brourd, thats only nice [3:58 PM]

Brourd: They use a different, more expensive, form of sequencing. [3:58 PM]

Eli Fisker: Machine - Cloud lab is what we run now [3:58 PM]

Eli Fisker: With many synthesis slots pr round [3:58 PM]

machinelves: is that a reference to running all synths in the same beaker? [3:58 PM]

Brourd: I'm sure we could make a C.L.O.U.D. acronym to describe it if you wish :) [3:59 PM]

Eli Fisker: Eterna classic lab, we mostly had 8 or 16 slots pr round [3:59 PM]

machinelves: ( hence the barcodes...?) [3:59 PM]

Brourd: Correct machinelves [3:59 PM]

Eli Fisker: I think the cloud part could be a referance to that they upload all the results to the interweb sky [3:59 PM]

machinelves: okay, thanks! :D [3:59 PM]

machinelves: hehe i see, the term cloud drives me nuts. leads people to believe the data is ethereal when really it is still sitting in a box somewhere [4:00 PM]

Eli Fisker: lol [4:00 PM]

Brourd: That is the reason why they called it cloud, instead of the "Lotsa RNA" lab. [4:00 PM]

jnicol: Cumulative Labs Organized Under Development [4:00 PM]

machinelves: haha [4:00 PM]

Eli Fisker: haha, John [4:00 PM]

machinelves: ahhhh thanks jnicol [4:00 PM]

Brourd: Nice [4:01 PM]

Eli Fisker: Good one [4:01 PM]

Eli Fisker: Quote [4:01 PM]

jnicol: ;) [4:01 PM]

Eli Fisker: :) [4:01 PM]

Brourd: Although, you need to use a cloud pun for the word cumulative. [4:01 PM]

machinelves: hehe [4:01 PM]

jnicol: cumulus clouds? [4:02 PM]

Eli Fisker: haha [4:02 PM]

Brourd: Sure, now mix that in there, and voilà. [4:03 PM]

jnicol: think I've hacked that acronym enough already :) [4:04 PM]

machinelves: sosounds like the cloud method is for testing en masse, and perhaps more specific methods could be used for finer grained followup on barcodeless designs [4:04 PM]

jnicol: barcode-less is expensive, which is why they went to clouds labs with the barcode identification [4:05 PM]

jnicol: much more throughput [4:05 PM]

machinelves: right right, hence the increased synth bandwidth [4:05 PM]

machinelves: exactly [4:06 PM]

jnicol: so why do you think they would want barcodeless labs? [4:06 PM]

machinelves: as a control to test against the barcode version of the same design [4:07 PM]

Eli Fisker: We already have the barcode less ones in Classic eterna [4:07 PM]

Brourd: So Eli,. is there any particular reasoning or what you hope to see in these new multiloop lab proposals? [4:08 PM]

Eli Fisker: I suggested making some copy labs with barcode [4:08 PM]

machinelves: i assume if any of these RNA bits are to be applied, the organism won't want the barcode part... or perhaps we'll just accept the barcode as part of the native chemistry of the design? [4:08 PM]

Brourd: @ machinelves - they did all that in 2012 :) [4:08 PM]

Eli Fisker: Brourd, I wish to fill in the data gap [4:08 PM]

Brourd: Ah [4:09 PM]

machinelves: right. i am looking for controlled experiments and the barcode seems like a wild card tossed in [4:09 PM]

Eli Fisker: And see if tendencies continue [4:09 PM]

Brourd: Tendencies? [4:09 PM]

jnicol: machine, we are not looking to put any of these designs in vitro, just looking for proper folds for now [4:09 PM]

Eli Fisker: Yes, there are multiloop tendencies [4:09 PM]

Brourd: @ machinelves - they worked with all that back in 2012 when they first created their protocol. [4:10 PM]

Eli Fisker: Late here. I should get to bed [4:10 PM]

machinelves: ya, just looking down the road. i doubt they'll consult us before moving to the inevitable next step, so i feel some responsibility in ensuring a clean protocol early on [4:10 PM]

machinelves: bye Eli! :) [4:10 PM]

Brourd: What would these multiloop tendencies be? [4:10 PM]

jnicol: *in vivo [4:10 PM]

machinelves: @brourd, kk thanks, i will look more into their protocol [4:10 PM]

Eli Fisker: Bye Machine  [4:10 PM]

Eli Fisker: Some of what I have already mentioned in forum and cloud lab notes [4:11 PM]

machinelves: i think maybe tendencies to bond or energetically affect other parts of the strand? [4:11 PM]

Brourd: hard to say Eli. I forget stuff easily :) [4:11 PM]

Brourd: http://arxiv.org/abs/1304.1072 here you go machinelves [4:12 PM]

machinelves: awesome, thanks brourd! [4:12 PM]

Brourd: Sure,  now all you have to do in exchange is pay $35 for me to get access to this paper I will link in a few seconds. [4:13 PM]

Brourd: Seems fair to me :) [4:13 PM]

Brourd: Actually, I have two papers, but you may be lucky enough to get them both! [4:13 PM]

machinelves: hehe i really wonder why they can't give us a journal portal if this is a university hosted project [4:13 PM]

jnicol: http://arxiv.org/ftp/arxiv/papers/1304/1304.1072.pdf [4:14 PM]

machinelves: but i don't know what the terms are [4:14 PM]

jnicol: for free [4:14 PM]

Brourd: Probably due to the fact we are not directly related to it. [4:14 PM]

machinelves: ah, well if that is the case then i guess they don't need all us lab mice pressing the levers that give them data... [4:14 PM]

Brourd: the free version of that should be right there in the top right, under PDF only [4:15 PM]

Brourd: on my link. [4:15 PM]

machinelves: thanks! [4:15 PM]

Brourd: http://www.ncbi.nlm.nih.gov/pubmed/12356306 here is one of the papers I need access to :) [4:17 PM]

machinelves: you and me both! :D [4:18 PM]

machinelves: is it true you can get alumni access to journals? i'll have to look into that... [4:18 PM]

Leila204: HI. Bye. Gotta Go [4:19 PM]

machinelves: cheers! [4:18 PM]

Brourd: the other is http://www.ncbi.nlm.nih.gov/pubmed/9454601 [4:19 PM]

Brourd: I have no clue machinelves :) [4:19 PM]

Leila204: XD Hi guys.  [4:19 PM]

machinelves: wow cool, thanks brourd! [4:19 PM]

Brourd: Both are somethign I want to look into, but I was unable to find a free copy :( [4:20 PM]

machinelves: ya i am glad to see a growing movement towards open publishing [4:20 PM]

Brourd: 10 years too late! :D [4:21 PM]

Brourd: or 15 for the one about wobble base pairs. [4:21 PM]

machinelves: part of the reason i left academia for private sector... seems like profs are more concerned with getting 'elite' publishing than accurate methodology / results, and it becomes a competition based environment that rewards not true results, but sensational results [4:21 PM]

machinelves: hence so many scientists faking or fudging results [4:22 PM]

Brourd: I always love a good sensation. [4:22 PM]

machinelves: and a host of other issues, one small consequence of which is the $$$ model of publishing data  [4:22 PM]

machinelves: lol [4:22 PM]

Brourd: "Cow travels faster than light"n pay $100 to find out how! [4:23 PM]

machinelves: hahahaha [4:22 PM]

machinelves: Population A applied neurosensory imagination receptors to bovine flight mechanism, producing a 73.08% possibility of lift-off [4:24 PM]

Brourd: I'll never be able to afford the cost of cow travel :( [4:24 PM]

Brourd: ha [4:24 PM]

machinelves: aww, don't worry, i hear they're working on Emu-based travel for us hoi polloi [4:24 PM]

Brourd: Thank you, but I prefer ostriches for my flightless bird travel. [4:25 PM]

machinelves: the feistier the better, eh? :D [4:26 PM]

Brourd: Indeed! [4:26 PM]

machinelves: hehe [4:26 PM]

machinelves: lol om nom nom. they echo the endless oooooooooooooo of wonder across the universe [4:29 PM]

Brourd: :D [4:29 PM]

Personal tools
Main page
Introduction to the Game