Scientists use ML to find an antibiotic able to kill superbugs in mice • Hacker News Board

Re: Scientists use ML to find an antibiotic able to kill superbugs in mice

#91

Post by perl4ever » Sun, Feb 23, 2020, 3:16 PM UTC

mnw21cam wrote: post #54

refurb wrote: post #32

Earlier quoted context omitted.

Super bugs are a mostly first world problem Not true at all. There are several known resistant strains that have come out of developing countries. Why? Antibiotic use can be rampant - in many countries you can buy them without a prescription.

I heard someone (in the correct area of expertise) describing how he had visited a developing country and tested the water in a river downstream of an antibiotic manufacturing plant, and found the concentration of a particular antibiotic to be approximately the same as what you would typically want to achieve in the blood of a patient. If that isn't going to drive resistance to that particular antibiotic, I don't kno…

People say things like this, but how many antibiotic compounds are in the soil naturally? Living creatures exist in an environment with countless bacteria and evolve to survive, and some of the compounds they produce are turned into human drugs. So I feel like there's something missing in the ordinary narrative about antibiotic resistance. Where did penicillin come from again?

Re: Scientists use ML to find an antibiotic able to kill superbugs in mice

#92

Post by Bartweiss » Mon, Feb 24, 2020, 8:15 PM UTC

SQueeeeeL wrote: post #87

conjectures wrote: post #86

Earlier quoted context omitted.

> People have been doing this exact thing for two decades at least but obviously with less computing power. Here's an analogy which shows how the cynicism here isn't actually useful wisdom: Lee Sedol lost at go. Kasparov lost at chess two decades before. The Kasparov loss was 'exactly the same thing albeit with less computing power'

Correct. I don't understand why this contradicts the argument. "Computers can learn complicated games with more computing power" "Compute power goes up" "Computer learns more complicated game" I mean, if you really think of AlphaGo on an extremely high level, it's just a really elaborate way to create and learn a dictionary of moves to take under different circumstances. Of course that's going to be completely depend…

I think the criticism is that it's not obvious whether success here was a function of improved performance, expanded throughput, expanded testing, or sheer luck.

Chess engines have clearly improved in both design and computing power over the years; doubling an engine's resources or pitting a new engine against an old one produces straightforwardly better play. But the drug-discovery technique in use here may not be "playing better" in terms of producing higher-quality predictions.

To extend the chess metaphor:

- Deep Fritz is a stronger player Deep Blue even with 4% as much computing power. This story does not appear to be an algorithmic breakthrough of that source.

- Deep Blue lost to Kasparov in 1996, then beat him in 1997 with double the computing power. That's a clear improvement in play, but not an improvement in efficiency. This story might represent such a change, modelling more prospective drugs to test higher-confidence candidates.

- If an AI that can only win 2% of games against humans plays 10 games, it has an 18% chance of beating someone. But over 100 games, it has an 87% chance of a win. This result might be a team with a larger testing budget claiming the 'first win' without any AI-side improvement.

- If a dozen grandmaster-level chess AIs play GMs, one of them will have to get the first win against a human. Labeling this result a 'breakthrough' in AI terms might be outright publication bias among equivalent projects.

As far as the drug, none of that really matters, except that efficiency improvements would have more potential to increase drug discovery. The drug itself is still useful, and the discovery is a proof of concept; in 1980 no possible computer would have beaten Kasparov. But this is being hailed as a breakthrough in AI in seriously questionable ways. The BBC article, for example, managed to imply that this specific project was novel and important for using neutral nets to produce a significant result.

Re: Scientists use ML to find an antibiotic able to kill superbugs in mice

#93

Post by Bartweiss » Mon, Feb 24, 2020, 8:35 PM UTC

redsymbol wrote: post #69

xzel wrote: post #18

Earlier quoted context omitted.

People have been doing this exact thing for two decades at least but obviously with less computing power. There's literally nothing new about the idea. The real trick is being incredibly lucky and finding something that actually works in humans after multiple trials. I'm sure you know this based on your comment and this isn't really directed at you (truly wish you best of luck, I really hope the computing power and s…

> There's literally nothing new about the idea. The real trick is being incredibly lucky and finding something that actually works in humans after multiple trials. I guess I'm not sure where the dismissiveness is coming from here. Are claiming this could have been trivially done before? If so, why didn't you or someone else do it already? Or are you claiming it's an uninteresting result that is not worthy of publicat…

This is a novel and important result in antibiotics. It's also a proof-of-concept for using ML to produce vital drugs with novel mechanisms, rather than incidental alterations or discoveries in noncompetitive spaces. It might be an incremental speedup or computing-power advance in ML drug discovery also, but it could equally just be the result of a lucky break or a particularly large lab-test budget. (In which case, "why didn't someone do it already?" is closer to asking why nobody else bothered to win the lottery.)

It's not a major theoretical advance in ML drug-discovery techniques or the first big step in ML drug discovery. It's certainly not the invention of ML drug discovery or neural nets as an ML technique, both things I've seen implied in news stories on this work.

This is attention-worthy, absolutely. (I'll leave "publication-worthy methodology" to experts.) But it's newsworthy on actual merits, as a drug breakthrough and a demonstration of an increasingly-important technique. So I share the frustration when lazy or confused reporting implies this is the same style of ML-theory breakthrough as CNNs, Transformers, or even neural nets themselves.