New EteRNA 2.0 proposal (retreat report)

Dear all,

EteRNA dev team had a long discussion about how EteRNA will change with the new high-throughput synthesis pipeline. Here are the details about the new EteRNA, which is temporarily called EteRNA 2.0 ~ Cloud Lab~ (If you have other ideas for the name, let us know!)

1. Roadmap based website

The new EteRNA 2.0 ~ Cloud Lab~ website will be focused around a concept of “roadmap.” The roadmap lists list of things you can clear and learn in an orderly fashion - so that the new players can simply follow the roadmap to become EteRNA experts. You may have seen similar concept in many other video games, or websites like “Khan Academy”

(A mock up of what a roadmap might look like)

Each node (or a stage) in the roadmap will have associated conditions and rewards which can be points, badges, and sometimes unlocking of game features.

The roadmap consists of 4 zones (only 2 showing up in the mock up)

i) Game & Puzzles : This is a part where you focus on learning basics of RNA folding, clearing challenges, making puzzles. This zone will embrace tutorials, challenges, and player puzzles.

ii) Cloud Lab :This is a part where you really get into player lab (or player projects) and rna synthesis. This zone comes after “Game & Puzzles”

This zone will have extensive nodes (stages) that will walk you through from participating in other players’ projects to creating your own project and succeeding on it.

iii) Paper & Publication : This is a zone which you can access when you have at least 1 “successful” player project. Stages in this zone will walk you through to create a scientific write up of what you discovered in your project. The write up will be published to “EteRNA Library” which will be an archive of all researches in EteRNA. Also, a few very good write ups will be encouraged to submit to scientific journals - (more details on cloud lab section)

iv) Bot Factory : This is still in discussion - but this part will focus on beating bots like Vienna, InfoRNA and RNASSD as we currently do in player puzzles. Furthermore this zone will focus on creating a new algorithm from “Player scripting interface” that can beat all other bots.

2. Cloud Lab

“Cloud Lab” is another term for “Player Projects” which will replace the entire lab section. Unlike current lab section, in which we developers propose one challenge at a time, players will propose their own lab projects and use up to 20,000 synthesis slots.

i) Project proposal

Any player who’s about half way through the Cloud Lab Zone will be able to propose a project. There are several things the player will have to prepare, but the most important thing is a “hypothesis” the player wishes to prove with the project.

A hypothesis should be a clear (and brief) statement that you want to test with your project. 1 good example would be “RNA won’t fold successfully if it has more than 90% GC pairs.” In the project you (and participants) will focus on proving that the hypothesis is true (or false)

In the early stage, we will restrict the number of active projects to 20 at a time. When a player propose a project, the project goes into “proposed projects” roster. All lab members will be able to vote on the proposed projects. If an active project finishes and a slot opens up, the project with the most votes at the time will go take the slot.

ii) Finishing a project

An active project can be finished in 2 circumstances. a) P.I (Principal Investigator, the project proposer) declares the project done. P.I moves on to “Paper and Publication Zone.” b) The project have been inactive for a certain amount of time

iii) Participating in a project & rewards

Any lab member can participate in a project by submitting a sequence. When a player submits a sequence to the project, he has to decide whether the sequence will “prove” or “disprove” the project hypothesis. For example, if a project’s hypthesis is “RNA won’t fold successfully if it has more than 90% GC pairs.”, you could create an RNA with 100% GC pairs and say that the rna will “prove” the hypothesis. Or, if you believe the 100% GC pairs RNA could fold successfully, you would say that the rna will “disprove” the hypothesis. You will get reward when your sequence is synthesized & the prediction is correct.

Players will also be able to comment on & “Like” other designs submitted in the projects.

iv) Synthesis

A single round of synthesis will take 2 weeks ~ 1 month time span. We are not perfectly clear how long it will take at this point. We will start by synthesizing 1,000 sequences per round, then eventually bump up to 20,000 sequences per round.

v) Synthesis slot allocation

As for the synthesis slot allocation, we decided to give full powers to players themselves.

Beginning of every round, all lab members will get approximately [total number of slots] / [number of lab members] synthesis slots in their hands.

A player can pick any sequence in active projects and choose to synthesize it (unless the sequence was already picked to be synthesized.) The player can pick his own sequence if he wants.

PIs of active projects, will get (possibly many) extra slots to spend on their projects.
These extra slots can only be spent on the PI’s project. Also, how these slots were used will be published. Technically, a PI could decide to spend the extra slots for only himself - but this will be known to all players. If offended players decide not to participate in the project anymore, the project will become inactive and get closed.

vi) Publishing to EteRNA library.

When a PI decides to finish his project, he’ll be able to enter “Paper & Publication” and formalize the outcome of the project as a scientific paper. The zone will guide the PI through step by step procedure to present the outcome of the project into a clear/concise writing.

Every finished paper will be published to EteRNA library, which will be an archive of all EteRNA researches. This will be a public archive, and the author of the paper will be fully credited.

Eventually, EteRNA team will pick 4 best papers out of EteRNA library and help the authors to further polish the papers and publish to scientific journals. We hope to do this in 1 year from now. We’ll also encourage 4 authors to help other players in polishing their papers and researches - the goal is to open the opportunity of publication to every member of EteRNA cloud lab.

3. EteRNA new website structre

(How pages in the new websites are connected to each other.)

If you have suggestions or questions about the new EteRNA, please let us know!

EteRNA team

For the most part this looks great. Thanks for all your efforts! There is one area that concerns me, and that is the allocation of synthesis slots, particularly when starting with just 1000.

"v) Synthesis slot allocation

“Beginning of every round, all lab members will get approximately [total number of slots] / [number of lab members] synthesis slots in their hands.”

This concerns me, because right now, there are far more qualified lab members than *active* lab participants. Right now over 2000 player have >10K points, unlocking the labs. And if starting with only 1000 slots, that means only the top 1000 players get 1 slot each per month. But how many of those would really get used??? Most players with lab access don’t use it.

For the current switch lab, 41 players cast a total of 190 votes on round 2. With 20 labs and 1000 slots that is 50 slots per month per lab. Each of those 41 players could use over *20* slots per month or 10 every 2 week lab cycle and still not fill the initial 1000 slots; and what happens when there are 20x more slots? I guess the hope is that more players will participate in labs if they are guaranteed a slot.

“PIs of active projects, will get (possibly many) extra slots to spend on their projects. These extra slots can only be spent on the PI’s project.”

This may help. If the balance of unused slots in each interval (2 weeks?) fall to the PIs and they get to rank (in advance?) the submissions to their projects, then they could get an average of up to 25 slots per lab every 2 weeks to fill in their submissions based on recommendations or “likes” by participating players but ultimately at their own discretion.

However, a lot of player projects don’t get anywhere near 25 submissions. Maybe that will change is there are only 20 active “cloud labs” at a time. But with 20000 slots, that would mean 1000 submissions per month per lab?! Once again, you must be hoping that more players will participate if guaranteed synthesis slots. And maybe they will.

Hi jandersonlee

We assume this will involve some trial & error. But we are planning to reserve a “redistribution” period for few days right after a round ends. Within this period, we’ll distribute unused slots to active lab PIs, and also active players. We even think maybe we can put these slots in an auction where the active players can buy the slots using game points. Or maybe these can be rewarded to active players (10 slots to active puzzle makers, 10 slots players who have been helping new people a lot, etc…)

Yes - we do hope and presume the active player will increase as the opportunity of synthesis increases. If that does not happen, then more slots for active players : ]

I like the idea of active players getting more. May be if you participate in a lab/project you get more guaranteed slots the next cycle? I’m not so fond of a few day “rush” to (re)allocate slots at the end. What if the players/PIs that get them don’t use them? Do the unused slots get wasted? I’d prefer a system where submissions are pre-ranked in some way and the distribution happens automatically.

I’m not at all sure about the idea of “buying/selling” slots. But then I’m not a big proponent of the free-market. :slight_smile:

Brourd: btw JL, you should mention [2:44 PM]
Brourd: the idea of a monthly registration [2:44 PM]
jandersonlee: ? [2:44 PM]
jandersonlee: Sign up for slots? [2:44 PM]
Brourd: in order to get lab slots [2:44 PM]
Brourd: you were wondering about the best way to determine when a player gets slots [2:44 PM]
jandersonlee: Register by the 15th for slots on the 30th? [2:45 PM]
Brourd: just, on a monthly basis, any player who signs in and does the necessary lab regitration will get slots [2:45 PM]
Brourd: no, monthly for both periods [2:45 PM]
Brourd: that way, players are not limited to a 15 day period [2:45 PM]
Brourd: to register [2:46 PM]
Brourd: so, if you register before the 15th, you are a part of round 1 and 2 [2:46 PM]
Brourd: if you register after the 15th, you will only be in round 2 [2:47 PM]
Brourd: registration can be easy [2:47 PM]
Brourd: something as simple as submitting a design should do [2:47 PM]
stevetclark: good idea brourd [2:48 PM]
Brourd: thanks [2:48 PM]
Brourd: btw, Hi steve [2:49 PM]
jandersonlee: Could do. I’d like to see the registration period expire at least 7 days before the labs are closed so people know how many slots they have to work with for a sufficient time to use them [2:49 PM]
Brourd: hmm, what happens if I work with African children for the first 8 days e=of every month? [2:49 PM]
Brourd: with no internet access [2:50 PM]
Brourd: no slots for me? [2:50 PM]
jandersonlee: :slight_smile: [2:50 PM]
Brourd: it has to be able to conform to those who may only be able to get on every two weeks basis [2:51 PM]
jandersonlee: I don’t know how often the will runn the labs. Last I heard it was once a month rather than every 2 weeks [2:52 PM]
Brourd: yeah [2:52 PM]
Brourd: several variables can change [2:53 PM]
jandersonlee: Two weeks to submit a “registration”, then two weeks to fill the allocated slots? [2:53 PM]
Brourd: well, otherwise, you are stuck filling 20,000 in one month [2:53 PM]
Brourd: same problem [2:53 PM]
jandersonlee: 41 active players 500 slots each?! [2:54 PM]
Brourd: hehe [2:54 PM]
Brourd: you assume of course [2:54 PM]
Brourd: no academics will take advantage of this [2:55 PM]
jandersonlee: I try not to make too many assumptions [2:55 PM]
Brourd: I’m sure we will see an increase in players from this [2:55 PM]
jandersonlee: But projecting the present onto the future is just that :stuck_out_tongue: [2:55 PM]
Brourd: :slight_smile: [2:56 PM]
jandersonlee: I’d just like a scheme that will give me *some* idea in advance how many slots I might have for a month (or whatever the inteval is) [2:57 PM]
Brourd: I wonder when the 1000 slots a month will start [2:57 PM]
jandersonlee: They were suggesting late summer as a possibility [2:57 PM]
Brourd: well, that is when you based it on previous labs :wink: [2:57 PM]
Brourd: for example, after the first few months [2:58 PM]
Brourd: then you will know an approximate amount [2:58 PM]
Brourd: after the first few months with 1000 slots [2:59 PM]
Brourd: how many players participated with that should help [2:59 PM]
jandersonlee: Mind if I post this chat segment as a comment on the EteRNA 2.0 topic? [2:59 PM]
Brourd: sure [3:00 PM]
Brourd: np to me [3:00 PM]

Are there any new details that you can share with us

Hi mat

you can keep track of our work at Eterna github repositary…