Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

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

statnews.com

Post reply on HN
Print view View original

21–30 of 93 posts

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

#21

Post by entee »

xzel wrote: post #18
entee wrote: post #2

This 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…

I share some of your skepticism, but I think this particular paper is worth a little less cynicism. It’s true that statistical analysis has always helped inform medicinal chemistry efforts. They used to be called QSAR ;) What’s different here is the scale of information that was processed. As you point out the computational power and therefore the algorithms we can apply has enabled different experiments which I very much hope will yield better and more abundant medicines. Papers like this are a sign of these methods bearing fruit!

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

#22

Post by cbhl »

koolba wrote: post #3

> 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…

Here is a prior HN submission that talks about the economics of new drug development for rare diseases:

https://news.ycombinator.com/item?id=19787367

tl;dr: a hypothetical drug with an expected profit of $547 million, adjusted for discounting and risk, has an expected value of negative $6 million.

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

#23

Post by mikorym »

ranDOMscripts wrote: post #17
DoreenMichele wrote: post #13

Earlier quoted context omitted.

I can't think of the last time I saw a headline on the front page of HN about the importance of solving open defecation. I see headlines about creating new treatments for super bugs pretty regularly. Open defecation costs $US260b globally https://www.news.com.au/world/breaking-news/open-defecation-... Price to pay: Antibiotic-resistant infections cost $2 billion a year http://www.cidrap.umn.edu/news-perspective/2018/…

That's probably because Bill & Melinda issued their grants a few years back. You can see a list of all of the high-tech outhouse makers here [0]. Many of the designs are out in the field for long-term testing. [0] https://stepsforsanitation.org/innovation-center/

I come from a country with serious sewage issues. Most of it is government operated sewage works that are just not maintained and were due for upgrade 10-20 years ago. The issue is also an order of magnitude worse in some areas than in others.

The Gates Foundation is not relevant here, in that problem field, in any meaningful way. It is an issue about operational structure and unit cost structure. Tax money gets spent on black Mercedes cars instead of what it is supposed to be spent on.

Also, it is not a technological issue at all. In terms of tech for sewage treatment, the path forward is clear. In terms of politics it is not.

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

#24

Post by entee »

chrisgd wrote: post #6
entee wrote: post #2

This 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…

Can you help me understand what is the difference between what you are doing and what people like Certara and Simulations Plus are doing? The simulation software market is really fascinating

Sure! To a first approximation they both provide software packages to pharma who then apply those to their efforts. In our case we are building an integrated system with a biochemical component and a computational component. We will then seek out molecules to difficult targets ourselves because we believe the integration gives us a competitive advantage.

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

#25

Post by entee »

naresh_xai wrote: post #20
entee wrote: post #2

This 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…

Exactly why we need causal reasoning/causal proof alongwith ML

In this case I’m not sure if the reasoning for what molecules work would make sense even if we had an “oracle” to explain it. Why a molecule works is a complicated interplay of fundamental physics and emergent properties. The explanation is likely not human interpretable without a ton of equations in the first place.

That said in broader medical cases of ML where particular symptoms and measurements drive a diagnosis for example, I tend to agree.

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

#26

Post by throwGuardian »

DoreenMichele wrote: post #9

...in mice. It would be nice if we spent as much time, money and attention on figuring out prevention. Inadequate hygiene infrastructure (like toilets) in developing areas is part of the problem here. But addressing that isn't as exciting to people as finding a cure for a super bug. If we really want to fix this, that needs to change.

Super bugs are a mostly first world problem, rampant in top tier hospitals providing cutting edge treatments, despite following safety precautions meant to deter secondary bacterial infections during or post hospitalization

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

#27

Post by throwGuardian »

What's becoming apparent is the fundamental nature of applied mathematics and ML/AI/DL to our future.

Our education system needs to adapt, include this as a mandatory part of a college education (BRIC countries are including this at the high school level). A degree in pure AI/ML without mastery over impactful problems and it's underlying science, is not the ideal future. Every chemist, biologist, ... should be proficient in ML/DL

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

#28

Post by DoreenMichele »

throwGuardian wrote: post #26
DoreenMichele wrote: post #9

...in mice. It would be nice if we spent as much time, money and attention on figuring out prevention. Inadequate hygiene infrastructure (like toilets) in developing areas is part of the problem here. But addressing that isn't as exciting to people as finding a cure for a super bug. If we really want to fix this, that needs to change.

Super bugs are a mostly first world problem, rampant in top tier hospitals providing cutting edge treatments, despite following safety precautions meant to deter secondary bacterial infections during or post hospitalization

Most developing countries don't actually keep adequate data on super bugs. So we don't really know that.

In 2014, the WHO released the first report on antimicrobial resistance, in which the WHO collected national data on nine bacterial infections/antibiotic combinations of greatest concern for global health.[5] The data revealed that out of 194 countries, only 129 provided data, of which only 22 had data on all nine infection-antibiotic resistance combinations deemed to be emerging global threats.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6380099/

Anyway, this is sort of off topic because I know the article was submitted to HN for the cool factor of using artificial intelligence or machine learning or whatever, not because people are really deeply concerned about antibiotic resistant infections. So idly saying "I feel like they are barking up the wrong tree and focusing on the wrong things if they want to solve this." isn't at all welcome as part of the conversation, as evidenced by the downvotes for a really mild, brief remark.

I wasn't trying to derail the discussion. I just read up some on things like antibiotic resistance because of my medical situation, so I happen to know a bit about the topic.

But I'm planning on stepping away from this discussion because it seems like a pointless waste of time that's just inviting negativity from a crowd that doesn't see my comments as anything intelligent or interesting.

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

#29

Post by thedance »

It's interesting here that the breakthrough seems to come from the computer search not suffering from epistemological block. It doesn't think, therefore it is also not bound by conventional patterns of thought, such as the bias a researcher may show when thinking that this known compound is not an antibiotic, when it is in fact an antibiotic compound.

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

#30

Post by ranDOMscripts »

mikorym wrote: post #23
ranDOMscripts wrote: post #17

Earlier quoted context omitted.

That's probably because Bill & Melinda issued their grants a few years back. You can see a list of all of the high-tech outhouse makers here [0]. Many of the designs are out in the field for long-term testing. [0] https://stepsforsanitation.org/innovation-center/

I come from a country with serious sewage issues. Most of it is government operated sewage works that are just not maintained and were due for upgrade 10-20 years ago. The issue is also an order of magnitude worse in some areas than in others. The Gates Foundation is not relevant here, in that problem field, in any meaningful way. It is an issue about operational structure and unit cost structure. Tax money gets spen…

It is my (albeit limited) understanding that the Gates grants are focused on providing toilets to areas where there previously were none and not on replacing crumbling infrastructure. They're trying to get facilities to the millions of people who have to poop in the bushes because that's all there is.
Post reply on HN