I saw news about this on The Guardian and it was the first time in a while where I was truly amazed by an ML-related discovery. This kind of thing is incredibly exciting and I hope we continue to see similar discoveries in ever more wider fields.
Scientists use ML to find an antibiotic able to kill superbugs in mice
61–70 of 93 posts
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#62Earlier 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…
I don't understand, they didn't find ONE they found NINE. That doesn't sound like luck.
Finding nine is more like luck then finding one. Note that just one of the nine (so far) is validated. If I tell you to roll a six on a dice and you do that’s pretty lucky. If I tell you to roll an even number (I.e., give you wider scope) and you do ... less amazing. They could have lab tested every compound in the set but it would have taken too long. The ML for them to 9. I’m not saying that this isn’t good work. Just addressing the statistical point re 1 v 9.
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#63Thank goodness it was "aided by machine learning". Otherwise, would anybody be reading this article ;)?
We're techies, not doctors. It's the computational solutions that interest us, only secondarily the underlying problems being solved. Thankfully, it doesn't matter a whit whether or not a bunch of programmers read about medical advancements on their lunch break.
There are a few scientists working in drug discovery here (I'm one of them), and quite a few from other domains from what I've read. We have philosophers that intervene regularly as well. Not just techies!
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#64Earlier 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…
I’m usually also skeptical of getting excited about successful animal trials for new medications, but with this I’m actually leaning more on the side of optimism. With cancer treatments and antivirals in mice, we’re not so much targeting the pathogen as targeting the host immune system in the hopes it ends up nerfing the intended target (tumor/virus/whatever). Given that the compound seems effective against C. Diffic…
Problem with many of these molecules that are found to work in mice, when put in human which is considerably more complex, say it goes thru the liver, the liver may break down most of it when circulating the blood stream. So now the drug needs to also resist that type of break down. Now multiply that by 10 to 100 of other types of reactions your body can generate from this.
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#65I'm not an expert in ML/Biology, but I wonder if doing so won't completely eliminate all of humankind's diseases, but shift the battle from one between humans and bacteria/viruses, to one where ML takes the place of humans by proxy (say we let this ML vs superbugs play out over centuries). I wonder to what direction evolutionary pressure in the face of ML would take bacteria. Perhaps a super smart bug.
Molecular biology is not that easy. Selecting a promising molecule is only a small fraction of the work. If you know the molecule you still have to be able to replicate and bring it to the place where it is required. So even if AI can support the process still a lot of plain old biochemistry is needed. AI is just one of many tools. And as you can read in the article a lot of trials are required by the regulatory authorities before the drug can be applied to humans.
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#66I'm not an expert in ML/Biology, but I wonder if doing so won't completely eliminate all of humankind's diseases, but shift the battle from one between humans and bacteria/viruses, to one where ML takes the place of humans by proxy (say we let this ML vs superbugs play out over centuries). I wonder to what direction evolutionary pressure in the face of ML would take bacteria. Perhaps a super smart bug.
Molecular biology is not that easy. Selecting a promising molecule is only a small fraction of the work. If you know the molecule you still have to be able to replicate and bring it to the place where it is required. So even if AI can support the process still a lot of plain old biochemistry is needed. AI is just one of many tools. And as you can read in the article a lot of trials are required by the regulatory auth…
True but it's interesting that once you have an oracle that given a molecule provides you with an antibiotic score, you can show it directly molecules that are easy to produce and or likely safe for humans. This should help significantly.
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#67Earlier quoted context omitted.
I’m usually also skeptical of getting excited about successful animal trials for new medications, but with this I’m actually leaning more on the side of optimism. With cancer treatments and antivirals in mice, we’re not so much targeting the pathogen as targeting the host immune system in the hopes it ends up nerfing the intended target (tumor/virus/whatever). Given that the compound seems effective against C. Diffic…
Problem with many of these molecules that are found to work in mice, when put in human which is considerably more complex, say it goes thru the liver, the liver may break down most of it when circulating the blood stream. So now the drug needs to also resist that type of break down. Now multiply that by 10 to 100 of other types of reactions your body can generate from this.
Are humans more complex than mice or just different? I've always assumed the latter, but I'm in no way an expert.
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#68Earlier quoted context omitted.
Molecular biology is not that easy. Selecting a promising molecule is only a small fraction of the work. If you know the molecule you still have to be able to replicate and bring it to the place where it is required. So even if AI can support the process still a lot of plain old biochemistry is needed. AI is just one of many tools. And as you can read in the article a lot of trials are required by the regulatory auth…
True but it's interesting that once you have an oracle that given a molecule provides you with an antibiotic score, you can show it directly molecules that are easy to produce and or likely safe for humans. This should help significantly.
It will not go without replication, trials and production, regardless what oracle you have. And keep in mind that the oracle has a quite significant error rate. You won't get an immediate solution, just proposals. And whether or how a drug works is also dependent on the environment and there are dynamic aspects and complex processes within a cell extending over space and time. It's much more complex than speech recognition or self-driving cars.
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#69This paper shows what can be done when you carefully run an ML program alongside a wet lab experimental program tailored to feed back into the ML program. The results end up far more interesting than some recent "ML aided drug discovery" papers in that they actually discovered a drug that functions very differently than known antibiotics (AB). Even though the structures that came out look AB-like, they work different…
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 publication or attention?
Re: Scientists use ML to find an antibiotic able to kill superbugs in mice
#70> That is an especially pressing challenge in the development of new antibiotics, because a lack of economic incentives has caused pharmaceutical companies to pull back from the search for badly needed treatments. Each year in the U.S., drug-resistant bacteria and fungi cause more than 2.8 million infections and 35,000 deaths, with more than a third of fatalities attributable to C. diff, according to the the Centers…
At least one disincentive is that if you do find an amazing new antibiotic effective against certain strains of antibiotic resistant bacteria, antibiotic stewardship means the medical community will try and use it only where necessary to slow any adaptation to the new drug. That makes your potential patient population much smaller.
Yeah, this seems game theoretically problematic. We clearly want a deep warchest of antibiotics, but new antibiotics won't sell in any meaningful quantity likely, or even hopefully, until after their IP rights expire. I don't see how you make money here.
Maybe the government should structure some new incentives for stocking the antibiotic warchest.