2019.09.26 Dev Chat
Omei: If I have my schedule right, there’s a meeting about to start? [3:59 PM]
Astromon: oh yes [3:59 PM]
Jieux: Dunno… I’m in the carpool lane waiting to pick up my kiddo. [3:59 PM]
hwaymentsteele: Hi all! Yup some of us are assembling here for a community chat on the riboswitch paper. [3:59 PM]
Jieux: if so, bad time for a meeting for me at least… unless it starts now and lasts 30 mintues. [3:59 PM]
Astromon: the paper [3:59 PM]
Jieux: this one https://docs.google.com/document/d/1eyd1Uca3QHObsogw73h8AWDbyIw23SqC6_lvyBvhlb4/edit?ts=5d81c67b<br></a># [4:00 PM]
Jieux: ? [4:00 PM]
Astromon: cool ! [4:00 PM]
Jennifer Pearl: hi [4:01 PM]
Astromon: my best score for the lab in this paper with results is a 77 [4:01 PM]
Astromon: and it didnt have a static stem like my other very low scoring ones [4:01 PM]
Astromon: i did some research for this chat< [4:01 PM]
Jennifer Pearl: A note about ED. If you are discussing Vienna 2 it is ensemble diversity if you are using NUPACK it is ensemble defect [4:02 PM]
Jieux: I read the paper I posted if that is the paper we are discussing. [4:02 PM]
Jieux: the conclusion is nicely worded . [4:02 PM]
ThomasG: Hello, this is Will Greenleaf (talking from my son’s account) [4:02 PM]
mgotrik: im here [4:02 PM]
rhiju: me too. hi everyone – thanks for joining us in community chat! [4:02 PM]
Jennifer Pearl: hi [4:02 PM]
Jieux: Heya rhiju… [4:02 PM]
nnyasin: hello [4:02 PM]
Astromon: Hi! [4:02 PM]
ThomasG: Ok, lets open the discussion of the paper! I hear people like the conclusion! yay! [4:03 PM]
ThomasG: Is there anything from the comments section of the google doc that could use a community-level discussion right now? [4:06 PM]
rhiju: i’m still reading the interchange between ‘Dru Orb’ and Jennifer Pearl – any of those two here? (Jennifer sent me an e-mail with some thoughts) [4:06 PM]
rhiju: Ah I see a comment from brourd [4:06 PM]
Jennifer Pearl: im here [4:06 PM]
jandersonlee: I only had typeos for the most part. [4:06 PM]
Jieux: Heya jandersonlee… saw your name in the edits :) [4:06 PM]
jandersonlee: And the examples. [4:07 PM]
Jieux: yea… well your author photo was out of focus :) [4:07 PM]
rhiju: great [4:07 PM]
rhiju: haha [4:07 PM]
jandersonlee: no prob. I’m luriking in this paper [4:07 PM]
rhiju: Brourd suggests that Figure 2 is busy [4:08 PM]
rhiju: (its in a comment near the end of the main text) [4:08 PM]
Jennifer Pearl: i agree [4:08 PM]
Jennifer Pearl: it took me a bit to figure out what was going on [4:08 PM]
jandersonlee: You tend to have busy diagrams. Don’t know if that’s a bio-science thing. [4:08 PM]
jandersonlee: It take longer to read most diagrams than the full text of the paper. [4:09 PM]
jandersonlee: s/read/comprehend/ [4:09 PM]
J3w1shGr4nm4: Barbosa [4:10 PM]
etaplin: this is ass [4:10 PM]
Barbosa100: ^ [4:10 PM]
nnyasin: I dont have friends :( [4:10 PM]
saperez: wow [4:11 PM]
Jieux: This is my favorite sentence in the paper "several participants independently began incorporating multiple input aptamers into their switch designs; thus ‘cheating’ the expected thermodynamic maximum activation ratio for a single-input switch. [4:10 PM]
saperez: wtf^? [4:11 PM]
etaplin: lol hannah has no points [4:11 PM]
saperez: wowwww vjay jay [4:11 PM]
Jieux: sounds like Fisker. [4:11 PM]
jandersonlee: Fisket et al [4:11 PM]
jandersonlee: (typos my specialty) [4:11 PM]
vgpurohit: pendeho [4:11 PM]
Jieux: I think I et aled a few of those… [4:11 PM]
saperez: jo* [4:12 PM]
vgpurohit: keonda cabroness [4:12 PM]
saperez: (-_-)/ [4:12 PM]
Barbosa100: Como andamos [4:13 PM]
nnyasin: cayate vj [4:13 PM]
vgpurohit: un poco loco [4:13 PM]
Barbosa100: aqui todo al 100 [4:13 PM]
ThomasG: I agree that fig 2 has a relatively large number of panels. Our options are likely to move elements to supplement or submit as is and likely move some to supplement later [4:13 PM]
Astromon: Single-input switches, revisited is what the paper went by? [4:13 PM]
rhiju: as a heads up we conversed with nando via e-mail – he has OK’ed the paper too [4:15 PM]
Astromon: i found it curious the paper would be done before the other rounds were synthesized. [4:15 PM]
rhiju: he also bowed out of being an explicit author as his contributions to this one were primarily technical (puzzle release) [4:15 PM]
rhiju: @astromon can you clarify? this sounds important [4:15 PM]
rhiju: this paper does not have ‘multi-input’ switches – incl. openTB, etc. – because we felt they should go in a separate paper [4:16 PM]
rhiju: already a lot in this one [4:16 PM]
Astromon: i could find one lab where we have results [4:16 PM]
Astromon: r1 single input switches [4:16 PM]
rhiju: can you post a link? [4:16 PM]
Astromon: i read where the paper is about other labs that havnt been syntasized yet [4:17 PM]
Astromon: https://eternagame.org/web/lab/7559749/ [4:18 PM]
rhiju: ty [4:19 PM]
Astromon: with all the labs we have done be it single input r2 or light-up labs ,i would just think those results would be important for a paper [4:19 PM]
rhiju: ah, yes good point. [4:19 PM]
hannah71013: are you guys in university? [4:19 PM]
jandersonlee: some, not all [4:19 PM]
Jieux: I’m in the grammar-school carpool lane… [4:19 PM]
Jennifer Pearl: lol [4:20 PM]
rhiju: @astromon we were not able to test those yet – we had serious problems with the imaging station platform; this is around the same time that we tried to make measurements on openTB round 4 [4:20 PM]
rhiju: but your point is a good one. [4:20 PM]
Astromon: Thanks [4:20 PM]
Jennifer Pearl: @rhiju Any thought on my comments? [4:21 PM]
Jennifer Pearl: or edits [4:21 PM]
Astromon: well obviously there was enough data in the r1 to do this paper so thats good! [4:22 PM]
rhiju: @astromon wait a minute we are revisiting. [4:23 PM]
rhiju: i think that in the writing of the paper we actually missed out on these reults – apparently we do have the data in hand (as “R107”) [4:23 PM]
Astromon: ah thats right [4:23 PM]
rhiju: thank you [4:24 PM]
mgotrik: isn’t R107 whats presented in Fig 3A now? [4:24 PM]
rhiju: @jennifer will (@thomasG) and I are still making our way through the comments [4:25 PM]
rhiju: @mgotrik we’re checking now (@hwaymentsteele is on it) [4:26 PM]
Jennifer Pearl: ok [4:26 PM]
hwaymentsteele: Figure 3A contains the tryptophan and theophylline switches from round 107. [4:27 PM]
hwaymentsteele: The data that’s missing is the FMN switches of round 107. Those ought to go in the progression of puzzles in Figure 2a. [4:27 PM]
rhiju: ah OK. So @astromon actually those data are in the paper [4:27 PM]
Jieux: Very Important: Did we get our labcoin for those labs? (just kidding) [4:28 PM]
Astromon: oh from the rapid rounds [4:28 PM]
rhiju: we called them ‘single-input switches revisited’ in Eterna and they correspond to your link. Perhaps we shouldn’t have called them ‘revisited’ because that’s where we tried to generalize from FMN inputs to tryptophan and theophylline [4:28 PM]
mgotrik: if i remember correctly, the arginine aptamer in R107 didn’t work as expected so we dropped it [4:28 PM]
rhiju: @jieux yea we did [4:28 PM]
Jieux: heh [4:28 PM]
rhiju: @jieux we haven’t yet assigned lab coin from light-up sensor rounds, though @mgotrik @omei and i are working on it [4:29 PM]
mgotrik: by as expected - the aptamer didn’t work so we couldn’t get working switches from it [4:29 PM]
Jieux: OK… heh… Now we just have to change the ($) symbol to make me happy… [4:29 PM]
rhiju: @jennifer looking at the comment on static stems [4:29 PM]
rhiju: @jieux what symbol is better? [4:29 PM]
jandersonlee: USD$ [4:30 PM]
rhiju: @jennifer there’s agreement that we need to state explicitly that many of those stems are 5’/3’ combining. but there was already such a statement … “by promoting stable base pairings between the 5’ and 3’ regions or by incorporating…” [4:30 PM]
Jieux: Just a blank worthless gold circle like it used to be would be fine if you must equate points with $… don’t let me detract from paper talk though… [4:30 PM]
rhiju: now there are two such statements - a little later in the sentence i think you added “(including long stacks where the 3’ and 5’ ends meet)”. do we need both? [4:31 PM]
jandersonlee: I’m not clear on the meaning of that phrase [4:31 PM]
Jieux: Will the results from this paper be worked into soft constraints for future labs? Are we learning anything that can be utilized in this fashion? [4:31 PM]
Brourd: Hi Rhiju. Dru orb (or brourd) is my default Google account on my computer. I forgot to change accounts when replying to that comment chain (oops) [4:31 PM]
Jieux: I always thought your orb was a little Dru… [4:31 PM]
rhiju: @brourd yea i know, just explained that to others on the stanford end. [4:32 PM]
Jieux: The children are being set free from their cages… I must go… Please post this chat thread somewhere for folks who could not attend or had to leave early… [4:32 PM]
hwaymentsteele: @jieux one ongoing project in the lab is training a folding engine on riboswitch data to be better able to predict fold changes [4:32 PM]
rhiju: @lfp6 do you know how to save this chatlog and post, as jieux suggested? [4:33 PM]
Jieux: thanks hwaymentsteele… I’m very intrested in bot training… but must go… [4:33 PM]
Jieux waves [4:33 PM]
J3w1shGr4nm4: Adios! [4:34 PM]
rhiju: see ya jieux [4:34 PM]
Jennifer Pearl: on a side note…im seeing some cool things with NUPACK and investigating the ensemble [4:34 PM]
Brourd: I believe the original language of the paper for that statement was fine (5’ and 3’ ends have to be base paired together, and something about hairpins). While it may be that there is some length correlation of ‘static stems’ (tbc) [4:35 PM]
Brourd: I believe that there was a line in the paper that goes on to say that more detailed features would be mentioned in the supplemental [4:36 PM]
ThomasG: just to be clear, the supplement appears after the main text and figures on the google document [4:37 PM]
Jennifer Pearl: I think I have provided ample data demonstrating the long stems work best but that ius besides the point. The point it that I “Purposfully” only had SARA submit designs with Long static stems for the labs [4:38 PM]
Brourd: Right. I think that would be a more appropriate place to disseminate more specific (or weakly supported) player design features [4:38 PM]
rhiju: @brourd are we missing details in the supplemental? [4:38 PM]
Brourd: I don’t think so? [4:38 PM]
LFP6: (Peeking in - I’ll make sure to post a chat log on the wiki par normal) [4:38 PM]
LFP6: *per [4:39 PM]
Jennifer Pearl: I have ran every designs for the labs through DPAT and showed the behaviour [4:39 PM]
Jennifer Pearl: i have nto done a strict study i agree [4:39 PM]
rhiju: @brourd you’re right. just put a comment in to make sure we get static stems back in [4:39 PM]
Jennifer Pearl: it shoudl have said static stems the whole time… [4:39 PM]
Jennifer Pearl: my whoel point is that if the paper is stating what players did to predict the good designs then the long static stems should be put in . Otherwise if the paper is to describe only proven ideas then just saying static stems is good. [4:41 PM]
Jennifer Pearl: that is the main point I have made [4:41 PM]
Jennifer Pearl: and continue to make [4:42 PM]
jandersonlee: Point taken. The idea of using static stems seemed fairly universal to many players, the length is something you investigated. [4:42 PM]
rhiju: @jennifer ok thanks, makes sense. we should probably stick to points that we can ‘back up’, or we will have trouble with reviewers (and readers) [4:43 PM]
Brourd: It’s not my place to speculate more on the intent. I think the original language was fine, as it did mention base pairing of the 5’ and 3’ ends, as well as the design of non-dynamic hairpins. [4:43 PM]
rhiju: i wonder though if you can write up the DPAT analysis in a separate short article [4:43 PM]
Jennifer Pearl: i have as white paper [4:43 PM]
Jennifer Pearl: https://www.dropbox.com/home/PublicShare?preview=Ideal+FMN+MS2+Riboswitches+FMN+Binding+Site+Stacks+and+Minimum+Pairing+Probabilities+Rev+1.pdf [4:44 PM]
Jennifer Pearl: i presented it at Eternacon [4:44 PM]
Jennifer Pearl: i can write something up new thought [4:45 PM]
Jennifer Pearl: thoguh [4:45 PM]
Jennifer Pearl: though [4:45 PM]
mgotrik: @Brourd - does fig 1H make sense to you? i.e. does the inset legend make it clear what is being shown? [4:45 PM]
rhiju: jennifer sounds good! we can’t see that link anymore. maybe join the eternapapers slack? [4:45 PM]
Brourd: I think it may be possible to write a manuscript on structural and sequence features of the data. There is a lot there, but I don’t know how much of it leads to strong correlative points. [4:45 PM]
Brourd: 1H? [4:46 PM]
mgotrik: errm 2H* [4:46 PM]
Brourd: I haven’t read a paper on FACS, so the only thing I had no clue about was FCS and SSC [4:49 PM]
mgotrik: @rhiju - i [4:49 PM]
mgotrik: made a big comment on fig 2, reiterating some of my previous comments on it [4:49 PM]
Brourd: *FSC [4:50 PM]
rhiju: folks – i’ve got to head out (same excuse as jieux!) – many thanks for joining us and helping review this paper. [4:50 PM]
hwaymentsteele: fwiw @brourd, forward scattering and side scattering [4:50 PM]
mgotrik: forward scatter and side scatter, FYI - just two measures of “granularity” of whats being measured [4:50 PM]
Astromon: Thanks! [4:50 PM]
rhiju: pls leave comments in paper ideally by end of day today! [4:51 PM]
rhiju: see you soon! stay tuned for more community chats coming up! lots of stuff happening in terms of both science and site/game updates… [4:51 PM]
