2020.03.27 Dev Chat

From Eterna Wiki

Astromon: ‎heya dosoon! ‎[3:06 PM]
stevetclark: ‎Welcome dosoonkim ‎[3:07 PM]
Astromon: ‎ok ‎[3:07 PM]
dosoonkim: ‎Hi Astromon! hello steve ‎[3:07 PM]
stevetclark: ‎Any updates on funding? ‎[3:07 PM]
dosoonkim: ‎Ah, I think Rhiju is the most appropriate person to answer that question ‎[3:08 PM]
stevetclark: ‎Ok ‎[3:09 PM]
dosoonkim: ‎But in general, I think there is only good news in funding, which is to say that plenty of funding is coming in or will soon come in ‎[3:09 PM]
stevetclark: ‎Thats what I love to hear! ‎[3:09 PM]
wateronthemoon: ‎yay to that! ‎[3:10 PM]
D'Wydd: ‎really good to hear ‎[3:10 PM]
dosoonkim: ‎So a short update from the dev team: we have been busy at work planning experiments, identifying targets, launching labs/puzzles to do our part in addressing the COVID-19 pandemic, specifically working on mRNA design for vaccines ‎[3:10 PM]
dosoonkim: ‎I’m happy to answer questions about what we’ve been working on, get any feedback you have on the labs that have been launched, what you hope to see, etc. ‎[3:11 PM]
stevetclark: ‎Are we going to do any labs with riboswitches? ‎[3:11 PM]
dosoonkim: ‎Yes–that is in the works! Designing riboswitch labs is a bit more complex than the puzzles that have been launched thus far, but riboswitch design will definitely be a crucial part of the OpenVaccine campaign ‎[3:13 PM]
Astromon: ‎woohooo! ‎[3:14 PM]
Astromon: ‎the ribo switch labs were fun. (: ‎[3:15 PM]
stevetclark: ‎It is so exciting to participate in cutting edge research! ‎[3:15 PM]
spvincent: ‎Do you think the current voting mechanism is the best way of selecting solutions? ‎[3:15 PM]
dosoonkim: ‎Ah okay–so maybe an overview of what we have planned might be helpful. ‎[3:16 PM]
wateronthemoon: ‎Thank you, dosoonkim, and to Riuju and all the team. For so quickly responding. And also for being such a force in open science. ‎[3:16 PM]
dosoonkim: ‎So zooming way way out, the overall goal is to test mRNA designs. This is a really complex problem, so we are looking to tackle this in three prongs: 1) player designs, 2) computational design, 3) some randomized control designs. ‎[3:17 PM]
dosoonkim: ‎Focusing on the player designs, we have two priorities: 1) validate the experimental workflow and 2) design puzzles. ‎[3:17 PM]
DigitalEmbrace: ‎Do you have a partner for mRNA vaccine or will you wait until you have preclinical evidence? Or would clinical trial be run by Stanford School of Medicine? Or are we only working on specific elements for other vaccines? ‎[3:18 PM]
DigitalEmbrace: ‎Oops, I meant to delete that! ‎[3:18 PM]
dosoonkim: ‎The voting mechanism was used to quickly select a subset of sequences for validating the experimental workflow (once we work out the kinks testing ~5-10 designs, we can think about how to scale to 100s of designs). ‎[3:18 PM]
dosoonkim: ‎In future rounds, we have a high-throughput experiment in the works that will allow us to test hundreds to thousands of designs, and most likely will not need voting. ‎[3:18 PM]
spvincent: ‎ok, tx ‎[3:19 PM]
Astromon: ‎great! ‎[3:19 PM]
LFP6: ‎@DigitalEmbrace We have set up really exciting partnerships… but we can’t talk about the details, sadly ‎[3:19 PM]
tone: ‎excellent! ‎[3:20 PM]
wateronthemoon: ‎why can’t you tell who? ‎[3:20 PM]
dosoonkim: ‎@DigitalEmbrace, I think Rhiju would be the best to answer that, the dev team has mostly been laser-focused on setting up the science and getting that up and going ‎[3:20 PM]
stevetclark: ‎It would be cool if Moderna is involved ‎[3:20 PM]
LFP6: ‎(I don’t even know the details) ‎[3:20 PM]
wateronthemoon: ‎ok, all in due time ‎[3:20 PM]
LFP6: ‎There are also NDAs and such involved ‎[3:20 PM]
stevetclark: ‎I wonder if we will get any media coverage ‎[3:21 PM]
Astromon: ‎And a fine job you white coats are doing! ‎[3:21 PM]
LFP6: ‎We’re working on that @steve ‎[3:21 PM]
LFP6: ‎We’re reaching out to media outlets, putting together some specific material ‎[3:21 PM]
tone: ‎bravo, people in white lab coats ‎[3:21 PM]
stevetclark: ‎If they need to interview players I am down to do an interview ‎[3:21 PM]
LFP6: ‎And we’re actually going to be asking for the community’s help in some of that. News post will be coming out shortly ‎[3:21 PM]
LFP6: ‎(Working out the specific details right now) ‎[3:21 PM]
stevetclark: ‎My roommate works for a news station in Charlotte ‎[3:22 PM]
dosoonkim: ‎It would be safe to say that media coverage, funding, partnerships are all in the works! But for now I think it’s really important to iterate on our structure for labs/puzzles and get the science right. ‎[3:22 PM]
dosoonkim: ‎Does anyone have any feedback on the puzzles, voting mechanisms, design selection, timing, etc.? ‎[3:22 PM]
wateronthemoon: ‎yes. The science is bottom line. ‎[3:22 PM]
Astromon: ‎i think about the voting, worzeize i think got the most pairs of anyone, and did not get iselected. ‎[3:24 PM]
Astromon: ‎1thing* ‎[3:24 PM]
wateronthemoon: ‎Voting seems a bit skewed sometimes. Might be good if devs looked at designs and offered feedback. ‎[3:24 PM]
Brourd: ‎These two initial puzzles are acting as the tests for the experiments. Will the number of puzzles be expanded beyond this amount? ‎[3:24 PM]
spvincent: ‎I think it would be good to take a second look at the voting mechanism on those puzzles where it exists. ‎[3:24 PM]
dosoonkim: ‎Yes, my vision is that we will have a puzzle per target, so as the number of targets expand, there will be more and more puzzles added. ‎[3:25 PM]
dosoonkim: ‎ So perhaps players can develop expertise, and you specialize in epitope design, and others specialize in longer mRNAs, etc. ‎[3:25 PM]
Astromon: ‎nice ‎[3:25 PM]
Brourd: ‎So, I have feedback on the idea of multiple puzzles that I need to discuss with Omei and LFP6 then. ‎[3:25 PM]
tone: ‎interesting. I could be a specialist ‎[3:25 PM]
Astromon: ‎(: ‎[3:26 PM]
wateronthemoon: ‎everyone brings their own strengths and has something to offer. You never know when serendipity will happen. ‎[3:27 PM]
tone: ‎I think we have to live with voting. It is another tool to sort the submissions on. ‎[3:28 PM]
dosoonkim: ‎Yes @Brourd, or anyone on the dev team would happily provide feedback! We’re all in very close communication these days. ‎[3:28 PM]
dosoonkim: ‎@tone yes, and once we move to the high-throughput experiments, mechanisms like voting may become less and less significant/critical ‎[3:29 PM]
Astromon: ‎for the most part the voting is good, i think too ‎[3:29 PM]
wateronthemoon: ‎so more designs going to synthesis ‎[3:29 PM]
Astromon: ‎maybe we could polish it a bit ‎[3:29 PM]
DigitalEmbrace: ‎Is anyone considering working on a serology test to detect specific proteins humans make in response to CV infection? Could that be an RNA application? ‎[3:30 PM]
Zama: ‎Do you expect to have more labs coming online in the next week? ‎[3:30 PM]
stevetclark: ‎Are there plans to synthesize the rest of the first round designs? ‎[3:30 PM]
SharkCrazy7: ‎Maybe having a sortable number of base pair stat could be useful in this specific lab ‎[3:30 PM]
Astromon: ‎will there be any data from r1 that could be useful designning for r2 ‎[3:32 PM]
dosoonkim: ‎@DigitalEmbrace unrelated to Eterna, but there is already a paper out describing this. https://twitter.com/davidrliu/status/1243377007016251400 is a good summary by a rather prominent scientist ‎[3:32 PM]
wateronthemoon: ‎Stabilizing mRNA delivery for potential vaccine seems most important. What else is important that we can do? ‎[3:32 PM]
tone: ‎The second round … I’ll make more submissions. ‎[3:32 PM]
dosoonkim: ‎@Zama yes! Our goal is to have many puzzles/labs released very soon. Currently we have two open labs that are soliciting many, many designs but more are in the works ‎[3:33 PM]
tone: ‎wow, more labs ‎[3:33 PM]
Astromon: ‎bring em on! ‎[3:33 PM]
Francesco97: ‎Hi, sorry, someone know how to mark bases in Android ‎[3:34 PM]
Zama: ‎More labs this week- trying to budget my time a little-lol ‎[3:34 PM]
dosoonkim: ‎@stevetclark Yes, since there was not experimental feedback between the lightning round and round 2, it would be good to test the rest of the lightning round designs. Working with our DNA synthesis partner today to see what is possible! ‎[3:34 PM]
tone: ‎Lol, indeed. Schedule : stay home. Check. ‎[3:35 PM]
wateronthemoon: ‎haha tone ‎[3:35 PM]
dosoonkim: ‎@Zama if you hav extra time, you can spend that extra time on the current open labs! Those are hard problems to crack. ‎[3:35 PM]
D'Wydd: ‎way back when, the original vision seemed like everyone would be able to participate regardless of their knowledge surrounding RNA and now it seems like people are going to need specialized knowledge. Has the direction changed for inclusion ‎[3:35 PM]
D'Wydd: ‎of a wider audience in the game component ‎[3:36 PM]
D'Wydd: ‎would we like to see 100000 solvers ‎[3:36 PM]
dosoonkim: ‎@D’Wydd @LFP6 can speak to this further, but our goal is to increase inclusion by vamping up the tutorials ‎[3:37 PM]
tone: ‎It seems tricky and interesting that we have to be sooo loose … looking for players, interesting puzzles, deevelpeers (ha ha) ‎[3:37 PM]
dosoonkim: ‎I am going to have to sign off soon–any last minute questions/feedback? ‎[3:39 PM]
LFP6: ‎Yeah @D’Wydd we do want to increase accessibility and how many people are able to contribute ‎[3:39 PM]
D'Wydd: ‎how do we contribute more? like partnerships etc. ‎[3:39 PM]
LFP6: ‎Part of that will be doing things like pushing for mobile and availability in other languages ‎[3:39 PM]
tone: ‎… ‎[3:39 PM]
Astromon: ‎actually there are players that i was helping in the first level in progressions just a week ago, that is now making great lab designs ‎[3:39 PM]
LFP6: ‎Part of it is that our tutorials are honestly pretty bad ‎[3:40 PM]
LFP6: ‎For the short term, we’re going to work on replacing our current tutorials with the ones from NOVA ‎[3:40 PM]
LFP6: ‎In addition to creating new content specific to our COVID efforts ‎[3:40 PM]
LFP6: ‎In the longer term - say, 6-12 months instead of the next couple months - we are planning on completely overhauling the tutorials ‎[3:40 PM]
LFP6: ‎This is something that I personally care a lot about and I think is one of our biggest issues currently ‎[3:41 PM]
LFP6: ‎New players get super frustrated with the tutorials. They’re long. They’re hard. They don’t teach well. They also don’t teach a lot of the newer skills that are relevant ‎[3:41 PM]
LFP6: ‎That needs to change ‎[3:41 PM]
dosoonkim: ‎OK I am signing off–need to put together the DNA sequences from the lightning round for synthesis! This was delightful–thank you for the chat! ‎[3:42 PM]
LFP6: ‎If we addressed that, I think we’d be able to significantly increase the amount of active players ‎[3:42 PM]
LFP6: ‎*active and contributing ‎[3:42 PM]
wateronthemoon: ‎What about a “library” of relevant information, esp. to COVID, in one place. Right now info is spread out among too many places. ‎[3:42 PM]
tone: ‎Starters can get stuck on a few hard tutorials, all right. ‎[3:42 PM]
Astromon: ‎Thanks! ‎[3:42 PM]
SharkCrazy7: ‎Would it be possible to see in the review tab of the labs the total number of pairs of a design? ‎[3:42 PM]
D'Wydd: ‎thanks dosoonkim ‎[3:42 PM]
stevetclark: ‎When will the results fro. The lightning round be available? ‎[3:42 PM]
Astromon: ‎‎will there be any data from r1 that could be useful designning for r2 ‎ ‎[3:43 PM]
kindler: ‎As a new player working through the tutorials right now, I’m glad to hear you’re planning an overhaul! If I hadn’t read that just now I probably would have given up on Eterna–now I know these are hard for everyone, not just me! ‎[3:43 PM]
tone: ‎Good replay, Astro ‎[3:43 PM]
Brourd: ‎@sharkcrazy7 if you click the headings in the review tab it will sort that column (I think you asked about sorting base pairs earlier) ‎[3:43 PM]
LFP6: ‎@SharkCrazy Sadly probably not :( That’s a super sticky part of the code, and my bandwidth is super limited right now (thanks to school and onboarding 4 (!) new devs that will be helping us for the next little bit, doing strategic planning, ‎[3:44 PM]
LFP6: ‎and any critical work for getting labs launched) ‎[3:44 PM]
D'Wydd: ‎awesome ‎[3:44 PM]
Astromon: ‎we could ask players to put data into the name ‎[3:44 PM]
Astromon: ‎like i do ‎[3:44 PM]
LFP6: ‎@steve I think 3 weeks was the number I heard? Not 100% sure ‎[3:44 PM]
LFP6: ‎@kindler Yes, they’re most definitely hard for everyone ‎[3:45 PM]
LFP6: ‎It’s a real problem ‎[3:45 PM]
Astromon: ‎it would free up some of my trime :D ‎[3:45 PM]
LFP6: ‎@Astro We won’t have data returned from r1 by the time r2 closes ‎[3:45 PM]
tone: ‎very funny … free time ‎[3:45 PM]
Astromon: ‎ok ‎[3:45 PM]
SharkCrazy7: ‎@Brourd Thanks but i wanted to know for the total number of pairs which is more relevant in the current lab, not each one separatly :) @LFP6 ok np, i was just proposing, because making mods of the best designs could be tricky for 1000+ ‎[3:46 PM]
SharkCrazy7: ‎designs ‎[3:46 PM]
LFP6: ‎Yeah, it would be a good idea ‎[3:47 PM]
LFP6: ‎There are counts of AU, GC, and GU actually ‎[3:47 PM]
LFP6: ‎I think ‎[3:47 PM]
Brourd: ‎You could probably do something like sort by G-C base pairs, and then use the designs with the most G-C base pairs as a generic starting point. ‎[3:47 PM]
LFP6: ‎So at least you could do the map ‎[3:47 PM]
LFP6: ‎*math ‎[3:47 PM]
Astromon: ‎percentages ‎[3:47 PM]
Brourd: ‎Note: The %'s don’t add up to 100 ‎[3:48 PM]
LFP6: ‎Oh are there not actual pair counts? ‎[3:48 PM]
Brourd: ‎No. There are pair counts ‎[3:48 PM]
Astromon: ‎they want pair count ‎[3:48 PM]
Astromon: ‎quickly ‎[3:48 PM]
Astromon: ‎in a glamce ‎[3:48 PM]
Astromon: ‎instruct players to put data in name in the intro to the lab ‎[3:49 PM]
LFP6: ‎If someone wants to go submit a PR to EternaJS, I’m happy ro approve it :) ‎[3:49 PM]
LFP6: ‎*ro -> to ‎[3:49 PM]
Brourd: ‎Total count does not exist. You can sort by other pair counts, and there will be a weak correlation between the highest number of some base pairs with the highest total counts. ‎[3:49 PM]
LFP6: ‎We had to take down the repo temporarily but I’m just about to bring it back up ‎[3:49 PM]