So here (I think nature has no paywall): https://www.nature.com/articles/s41587-024-02175-6
Authors report on how they made vast progress using a minigame within a popular video game to get high numbers of data oints by game-playing people.
This is citizen science at its best, when people help science without even knowing and no expertise required.
@sonhouse
BLS crowdsources a multiple alignment task of 1 million 16S ribosomal RNA sequences obtained from human microbiome studies.
@Ponderable
Wouldn't that mean to get correlation data they would have to have DNA samples from the players as well as the game data?
@sonhouse saidNo they obtain snippets of RNA and then they match what they have and try to reconstruct the original.
@Ponderable
Wouldn't that mean to get correlation data they would have to have DNA samples from the players as well as the game data?
@Ponderable
Snippets of RNA from a million people? How do they do that? Snippets from other folks not in the study?
Oh, I get it now, it helps to actually read the article🙂 So the game is to put the snippets together by the game, now it makes sense. Kind of like having supercomputers on the cheap, eh.
@Ponderable
Wow, I didn't know it was that hard a problem. So how does that work, they get detailed images of the chemical formulas involved or what?
@sonhouse saidSo you have four letters which form a sequence. If you have several RNA-strands a few hundred bases long and snippets of some dozen bases long it is very difficult. Y
@Ponderable
Wow, I didn't know it was that hard a problem. So how does that work, they get detailed images of the chemical formulas involved or what?
ou need at least four to six matching letters on each end to reasonably assume that they fit. (It is more easy if you know start and end codons, but now we go into details.
@Ponderable
A huge job for sure. Wonder if and when quantum comps get real if that would do the job, and what about the top supercomps, now in the Exaflop region.