Wanted to share what I thought the interesting parts. From the university press release.
"To date, AI has been leveraged as a tool for predicting which molecules might have therapeutic potential, but this study used it to describe what researchers call “mechanism of action” (MOA) — or how drugs attack disease.
MOA studies, he says, are essential for drug development. They help scientists confirm safety, optimize dosage, make modifications to improve efficacy, and sometimes even uncover entirely new drug targets. They also help regulators determine whether or not a given drug candidate is suitable for use in humans... A thorough MOA study can take up to two years and cost around $2 million; however, using AI, his group did enterololin’s in just six months and for just $60,000.
Indeed, after his lab’s discovery of the new antibiotic, Stokes connected with colleagues at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) to see if any of their emerging machine learning platforms could help fast-track his upcoming MOA studies.
In just 100 seconds, he was given a prediction: his new drug attacked a microscopic protein complex called LolCDE, which is essential to the survival of certain bacteria.
“A lot of AI use in drug discovery has been about searching chemical space, identifying new molecules that might be active,” says Regina Barzilay, a professor in MIT’s School of Engineering and the developer of DiffDock, the AI model that made the prediction. “What we’re showing here is that AI can also provide mechanistic explanations, which are critical for moving a molecule through the development pipeline.”
> Indeed, after his lab’s discovery of the new antibiotic, Stokes connected with colleagues at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) to see if any of their emerging machine learning platforms could help fast-track his upcoming MOA studies.
It must be so cool to work at a university. You can just walk across campus to meet with experts and learn about or apply the cutting edge of any given field to solve whatever problem you're interested in.
Working in a university that has significant research facilities is awesome.
Even when you work in administration, learning opportunities abound and are easy to seize.
I’m too shy to just walk into a random lab and ask questions, but three times a year, my boss likes to organize a tour of a different research facility and I really appreciate that.
You can email them and they are usually quite happy to talk.
Of course not when they get a media storm like these people. But I regularly correspond with experts in adjacent fields who have interesting papers put out.
In my sample of 1 with cold email to a researcher was that they are enthusiastic when someone has read their paper and ask relevant questions.
I don't remember the paper subject neither the researcher name (more than 20yr ago) but I remember that she was an ornitologist, the subject was quite niche and the response I received to my questions was longer than the article that prompted me to asks them.
Can confirm that most of the times when reproducing/implementing a paper or trying to extend it to another field, researchers are pretty OK (some very enthusiastic) to chat over email about it. As long as you've actually read the paper(s) or read the code (if any), and there's no expected free-work...
I sometimes get unpublished artefacts (matlab/python/fortran code, data samples) just by... asking nicely, showing interest. And I'm not even in academia or a lab.
The Reverse Gell-Mann Amnesia Effect-- one vastly underestimates the work it took to reach a conclusion in a book they haven't finished reading. Then, without reading another page, they assume everything in the entire bookshelf is at the same shallow level as their own misapprehension of the book they didn't finish.
I rankly speculate: for the set of low-effort comments on HN, there are more Reverse Gell-Manns than there are Gell-Manns.
There are 1000s of state and community colleges. Not everyone is doing groundbreaking research. Some places, they just teach, which is fine. Additionally, I have no idea what you're trying to say.
Because that is what the general public believes AI means, and Open AI say they are building thinking machines with it, and this headline says ”predicted”.
We knew that LolCDE was a vulnerability to e coli since well before 2016 and knew inhibitors of the complex, globomycin being one of them, which they knew about since 1978
> A thorough MOA study can take up to two years and cost around $2 million; however, using AI, his group did enterololin’s in just six months and for just $60,000.
Beautiful, finally something for ai/machine learning that is not a coding autocomplete or image generation usage.
It would be very interesting to keep track of this area for the next 10 years, between alpha fold for protein folding and this to predict how it will behave, how cost is reduced and trials get fast tracked
> We thus frame molecular docking as a generative modeling problem—given a ligand and target protein structure, we learn a distribution over ligand poses.
I just hope work on these very valid use cases doesn’t get negatively impacted when the AI bubble inevitably bursts.
In light of the meta context that this article reinforces the view that ai can replace researchers job I found this part of the artcile very true to how I use AI tools at work.
"Stokes stresses that while the prediction was intriguing, it was just that — a prediction. He would still have to conduct traditional MOA studies in the lab.
“Currently, we can’t just assume that these AI models are totally right, but the notion that it could be right took the guesswork out of our next steps,”...so his team, led in large part by McMaster graduate student Denise Catacutan, began investigating enterololin’s MOA, using MIT’s prediction as a starting point.
Within just a few months, it became clear that the AI was in fact right.
“We did all of our standard MOA workup to validate the prediction — to see if the experiments would back-up the AI, and they did,” says Catacutan, a PhD candidate in the Stokes Lab. “Doing it this way shaved a year-and-a-half off of our normal timeline.”
Same. I note in-advance that I'm not sure whether you yourself are referring to use of LLM tools in your research or rather the results of your own domain specific application of deep learning etc -- here, I assume the former.
I feel like the common refrain of most LLM success stories over the past year is that these tools are of significantly greater help to specialists with "skin in the game", so to speak, than they are to complete amateurs. I think a lot of complaints about hallucinations reflect the experience of people who aren't working at the edge of a field where they've read all the existing literature and there simply aren't other places to turn for further leads. At the frontier, moreover, the probability that there exists a paper or book that covers the exact combination of topics that interests you is actually rather low; peer discussions are terrific, but everyone is time-starved.
Thus I find the synthetic ability of LLMs to tie together one's own field of focus with those you've never thought about or are less familiar with to be of incomparable utility. On top of that, the ability to help formulate potential hypotheses and leads -- where of course you the researcher are ultimately going to carry out the investigation or, in the best case, attempt to replicate results. Conversely, when I'm uncertain of my own conclusions, I often find myself feeding the best LLM I have access to the data I reasoned from to see whether it independently gets to the same place. I'm not concerned about hallucinations because I know there's nobody but me ultimately responsible for error -- and, at the fringe of knowledge, even a total fabrication can inspire a new (correct) approach to the matter at hand.
I think if I had to succinctly describe my own experience it would be that I never get stuck any more for days, weeks, months without even a hint of where to turn next.
Related, there's an ancient Palantir blog post (2010!) that always stuck in my memory about a chess tournament that allowed computers, grandmasters, amateurs and any combination of the above to enter [0]. At that time, the winning combination turned out to be amateurs with the best workflow for interfacing with machine. The moral of the story is probably still true (workflow is everything), but I think these new tools for the first time are really biased towards experts, i.e. the best workflow now is no longer "content neutral" but always emerges from a particular domain.
While I agree, one must be careful anyway. I'm ignorant in most fields, reasonably good at two, and quite good (but far from excellent) in one. So while there is a lot to learn in the former, when it comes to the latter, all LLMs, including SOTA models, give me a very high percentage of answers that are misleading, wrong, dangerously incomplete, only superficially correct, amalgamate of correct and incorrect bits etc. Knowing this first hand, repeatedly, on hundreds and hundreds of issues, I basically built a deep, methodological distrust towards LLMs answers. In the end, I assume the answer to be wrong, but I look for verifiable hints that could lead me in the right direction. This is my default working mode in my niche.
To be clear, what the researchers & AI here have discovered is not a treatment for IBD per se.
Rather, the gut of some people, especially people with IBD and people who have received broad spectrum antibiotics, can be colonized by enterobacter species. These are bacteria (including some kinds of E. Coli) that are resistant to broad spectrum antibiotics, and this overgrowth is not good for gut health. The researchers have discovered a compound that appears to fight these enterobacter species without destroying the larger gut microbiome. This could help people (especially people with IBD) whose gut has been taken over by this kind of bacteria get back to a more normal gut microbiome, although only mouse studies have been done so far.
“This new drug is a really promising treatment candidate for the millions of patients living with IBD... We currently have no cure for these conditions, so developing something that might meaningfully alleviate symptoms could help people experience a much higher quality of life.”
We don't know. It's probably a combo of genetic factors and bacteria. It's at the very least complicated and multifaceted, because you can't just take a culture or biopsy of inflamed skin and say "A ha! This is clearly different!"
All the most effective treatments try to turn off parts of the immune system, and even though have minor success with some patients going through multiple different immunosuppressants to find the right one, or even cocktail, that adequately manages the disease.
It's also the case that there is a bewildering variety of things that get sort of lumped together as "IBD." Crohn's and ulcerative colitis are two of them, but there's no particular reason to assume inflammatory bowel diseases all have the same set of causes. Pretty much all of them are made worse by an e. coli infection, though, so a drug that can target just those bacteria is helpful!
During my own IBD journey, I've managed to stump the heck out of two different teams of GIs. I had been diagnosed with UC by biopsy during colonoscopy, and then at my last colonoscopy, despite not having been on medication for more than two years, they determined not only that I don't have it now, but that I never did. They told me "remission" would look different from "this bowel has never had IBD." But they also insisted I had not been misdiagnosed.
And yet they told me with a straight face that it is incurable. I had it in the past, confirmed by pathology. I don't have it now. And it's incurable. I give up.
In the end, I don't care enough to fight them about the contradiction, because the part I most care about is the "I don't have it now" part, and we're all in agreement on that.
(Note for any who are interested: I stopped medication after successfully reducing my inflammation markers within normal limits by eating the exact same thing for every single meal for 20 months with no cheating of any kind. They told me that shouldn't have been possible either, but it worked. And yes, it was as miserable as it sounds, but less miserable than living with UC.)
Sounds like you first decimated the bacteria with antibiotics, and then made it heavily disadvantaged by eating something that preferentially feeds other gut bacteria. What was the food, out of curiosity?
Actually when I medicated it, it was with mesalamine, not an antibiotic. I didn't do anything to attack the bacteria.
The meal consisted of well-boiled meat (pork/beef/chicken together, slow cooked, overnight), and carrots, zucchini, and butternut squash cooked in its juice until it mostly wasn't soup anymore.
The meal consisted of well-boiled meat (pork/beef/chicken together, slow cooked, overnight), and carrots, zucchini, and butternut squash cooked in its juice until it mostly wasn't soup anymore.
I wanted it to be soft/easy to digest, but also nutritious. This was at least a wide enough variety of nutrients that I didn't feel miserable (just bored).
CRP and fecal calprotectin. The meal consisted of well-boiled meat (pork/beef/chicken together, slow cooked, overnight), and carrots, zucchini, and butternut squash cooked in its juice until it mostly wasn't soup anymore. Picked things that were very soft/easy to digest, with at least some nutritional value, because it didn't bother me. I just wanted to buy time for things to calm down.
It was meant to be an elimination diet with reintroduction, but every reintroduction attempt failed miserably for 20 months. Then, suddenly, it was fine.
I still mostly eat food I would recognize in its ingredient form instead of highly processed stuff, but if everybody's going out for pizza, I can have a couple slices and be fine. I just can't do that all the time.
Not the person you're replying to, but my own IBD went away completely after I did keto for an extended period. This was a decade ago and it hasn't come back even as my diet returned to normal.
In my n=1 experience, it seems maybe significant dietary changes can perturb the gut ecosystem out of whatever state corresponds to IBD.
I had a similar experience, i.e. strict keto for 9 months and fully came off meds with no signs of UC or asthma. UC came back after around 2 years and I don't remember how long before I had to start asthma meds again. I can no longer do keto though, each time I try I get palpitations now after around a week.
Yeah my asthma went into remission as well, though this was also a period of my life when I was very physically active and in the best shape of my life, so it's hard to say what was the cause there.
They sometimes can, but most properly gut-adapted microbes only have their relative prevalence reduced by diet changes, by and large people unfortunately tend to snap back once diet is discontinued.
Ideally, you'd combine that with doing a stool transplant first, since it's the main thing (other than just antibiotics) that causes permanent compositional changes.
I'm not a scientist on any level, so my theory only went as far as "stop irritating the thing and let it heal itself."
When they started talking about putting me on immune-suppressant drugs during a pandemic, I thought that didn't sound like a very good idea, but maybe my body could sort itself out if I gave it the opportunity.
Not a terribly sophisticated take, but it works for things like not popping blisters or picking at scabs, so it seemed worth a shot.
i do the same, eating same thing for 2 years. While i don't have big IBD symptoms, i cannot now introduce new food, every time i try to introduce even very very small doses, i get a strange disproportionate reaction of my gut.
Gosh, I'm sorry. It does suck. For me, it was like that until it suddenly wasn't anymore. Nothing obviously changed. I kept "testing" reintroducing very small amounts of other food to try to end the elimination diet, like a bite every two months or something, and after 20 months, suddenly it went fine.
I say that just in hopes of encouraging you that healing might be right around the corner and you just might not know it yet. I certainly didn't know it was about to be over when it ended.
As someone with IBD who is on their fifth trip to the toilet since midnight (BST) I'd love to get the recipe. I've got a few months of steroids ahead and anything more long term and high impact is of enormous interest.
The meal consisted of well-boiled meat (pork/beef/chicken together, slow cooked, overnight), and carrots, zucchini, and butternut squash cooked in its juice in a pan until it mostly wasn't soup anymore. Salt to taste, but it's never gonna be delicious. It's tolerable. It's even fairly enjoyable once or twice. It's not a bad meal. But boy is it a tedious one after awhile.
Doctors are the only people I’ve encountered who are ready to inform you of their poorly supported contradictory conclusions with full confidence, and are fully ready to meet any pushback with gaslighting or the dreaded “difficult patient” label.
They approach the phenomenon of dropping trust and respect for their profession in much the same way.
It’s really frustrating. I don’t get why they feel entitled to acting this way when no one else does.
After making you wait 40 minutes later than when your appointment was supposed to start, I’ve looked at your chart for about 2 minutes and have spoken to you 60 seconds I’ve confidently diagnosed you with X. Here’s a prescription, let’s see you back in 6 months.
My favorite is the "bless your heart" head tilt that strongly implies you're just too damn stupid to follow what they're saying, when it's obvious on its face that they're contradicting themselves.
"I was correctly diagnosed with it before."
"Correct."
"I do not have it now."
"Correct."
"It is not 'in remission.' I just do not have the disease."
"Correect."
"But it isn't curable."
"Correct."
"So if it wasn't 'cured,' what did happen to it?"
{head tilt} "You tell me. Follow it through. You used to have it. You do not have it anymore. It's not in remission. But it also hasn't been cured. What does that leave?"
"#$%^! I don't know! That's why I'm here, asking you! As far as I know, all it leaves is contradiction and impossibility!"
"Now, now. There's no need to be difficult."
(This is almost word for word the last conversation I had with my GI. I'm not exaggerating. I'm not paraphrasing. I went through it exactly like this, and she responded to me exactly like this.)
My trick is to learn the terminology/jargon of the specialists by buying meds students used textbooks. It completely changes the conversation dynamics even if initially it appears to slightly confuse them.
Sometimes it angers them but then it's a clear signal that you must see someone else ASAP. When we grilled my wife first oncologist on his protocol, he broke down and said that he was not up to date on the latest research. We requested someone else, he was a much better fit and most importantly she is still alive, her metastasis disappeared and the latest scans and bloodworks results are still NED (no evidence of disease).
Yes, that was one of the inflammatory markers they checked. At diagnosis, I was in the mid-400s. After medication, I was in the low-200s. After 20 months of this elimination diet, I was in the 20s.
Multiple doctors told me it would not be possible to decrease the score without medication, one of them even while holding the results in his hands proving that I had done it.
Not directly since we all have it. It’s a bit confusing from the article but basically they discovered an antibiotic which can target a narrower spectrum than before which will minimize messing up your natural flora and allow drug resistant enterobacter (e coli is a main species, but not the only one) to be over represented. There is an association with ibd and. microbiomes that have e coli over-represented. And there are some subtypes of e coli that appear to be associated, see:
not exactly but E. coli and other Enterobacteriaceae can actually use the inflammation to their advantage
Inflammation release nitrates, which Enterobacteriaceae can use as a replacement for oxygen as a terminal electron acceptor
They can also encroach more easily on the protective mucus layers, which are thinner and more porous during inflammatory conditions (which itself may be a result of a messed up microbiota, which is why broad-spectrum antibiotics are not a solution)
These Enterobacteriaceae blooms in turn can cause inflammation, makes remission harder
There has been some success in reducing inflammation levels in IBD by blocking some of the binding factors that E. coli uses to attach to the epithelium/mucus
> Currently, we can’t just assume that these AI models are totally right,
Why could we ever assume that?
> but the notion that it could be right took the guesswork out of our next steps,
Devils advocate here. Couldn't this just be a severe case of confirmation bias? You take 100 such cases, ask AI "how does it work?" and in 99 of those, the answer is somewhere on the spectrum between "total nonsense" and "clever formulation but wrong". One turns out to be right. That's the on we are seeing here, getting confirmed in the lab. That doesn't actually mean AI reduced the time by 75%.
A broken clock is also correct twice a day. We wouldn't say we have invented a clock that works without energy, sure it's wrong sometimes, but when it's correct, it's awesome! No, it's just a broken clock that's wrong most of the time.
I would also love to see that with "generative AI" we have discovered some helpful magic, but as long as we are not honest about those details (which would include publishing and owning up to mishaps), this is all just riding a hype train.
I think this perspective overlooks how human expertise actually works. Humans in cutting-edge research also get things wrong a lot: most hypotheses fail, most experiments don't pan out, and most novel approaches lead nowhere.
When we celebrate a scientist who makes a breakthrough, we're not crediting them for being right 100% of the time. We're recognizing that they were right more often than random chance and earlier in the process than would otherwise occur.
A researcher (or AI) who can identify promising directions at a 2/99 or 3/99 rate instead of 1/99 is genuinely valuable – they're effectively doubling or tripling the efficiency of the discovery process.
Imagine if AI can test theories in under 100 seconds AND is slightly better out of 99 tries at getting things right. Beats the human out of the water.
This is exactly how I use AI now daily: Provide it with relevant context and help me to troubleshoot some software issue. I found often it casts a wider net and provides good leads to follow up. It balances out my own bias and overall speeds up discovery of the root cause. Admittedly some benefit comes from having to explain the AI the context - half the problems are gone once one can explain them clearly.
Machine-learning has been used in scientific research for a decade?
Is there something new?
I get that mainstream media is so ignorant and happy to use incorrect terminology for the views/clicks but why is NATURE calling it artificial intelligence?
Here is the original study published in nature microbiology.
https://www.nature.com/articles/s41564-025-02142-0
Wanted to share what I thought the interesting parts. From the university press release.
"To date, AI has been leveraged as a tool for predicting which molecules might have therapeutic potential, but this study used it to describe what researchers call “mechanism of action” (MOA) — or how drugs attack disease.
MOA studies, he says, are essential for drug development. They help scientists confirm safety, optimize dosage, make modifications to improve efficacy, and sometimes even uncover entirely new drug targets. They also help regulators determine whether or not a given drug candidate is suitable for use in humans... A thorough MOA study can take up to two years and cost around $2 million; however, using AI, his group did enterololin’s in just six months and for just $60,000.
Indeed, after his lab’s discovery of the new antibiotic, Stokes connected with colleagues at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) to see if any of their emerging machine learning platforms could help fast-track his upcoming MOA studies.
In just 100 seconds, he was given a prediction: his new drug attacked a microscopic protein complex called LolCDE, which is essential to the survival of certain bacteria.
“A lot of AI use in drug discovery has been about searching chemical space, identifying new molecules that might be active,” says Regina Barzilay, a professor in MIT’s School of Engineering and the developer of DiffDock, the AI model that made the prediction. “What we’re showing here is that AI can also provide mechanistic explanations, which are critical for moving a molecule through the development pipeline.”
> Indeed, after his lab’s discovery of the new antibiotic, Stokes connected with colleagues at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) to see if any of their emerging machine learning platforms could help fast-track his upcoming MOA studies.
It must be so cool to work at a university. You can just walk across campus to meet with experts and learn about or apply the cutting edge of any given field to solve whatever problem you're interested in.
Working in a university that has significant research facilities is awesome.
Even when you work in administration, learning opportunities abound and are easy to seize.
I’m too shy to just walk into a random lab and ask questions, but three times a year, my boss likes to organize a tour of a different research facility and I really appreciate that.
Can confirm it is often very cool to be able to do this.
That's only if there are experts there. The average college is not really what you're thinking it is.
You can email them and they are usually quite happy to talk.
Of course not when they get a media storm like these people. But I regularly correspond with experts in adjacent fields who have interesting papers put out.
In my sample of 1 with cold email to a researcher was that they are enthusiastic when someone has read their paper and ask relevant questions.
I don't remember the paper subject neither the researcher name (more than 20yr ago) but I remember that she was an ornitologist, the subject was quite niche and the response I received to my questions was longer than the article that prompted me to asks them.
Can confirm that most of the times when reproducing/implementing a paper or trying to extend it to another field, researchers are pretty OK (some very enthusiastic) to chat over email about it. As long as you've actually read the paper(s) or read the code (if any), and there's no expected free-work...
I sometimes get unpublished artefacts (matlab/python/fortran code, data samples) just by... asking nicely, showing interest. And I'm not even in academia or a lab.
The Reverse Gell-Mann Amnesia Effect-- one vastly underestimates the work it took to reach a conclusion in a book they haven't finished reading. Then, without reading another page, they assume everything in the entire bookshelf is at the same shallow level as their own misapprehension of the book they didn't finish.
I rankly speculate: for the set of low-effort comments on HN, there are more Reverse Gell-Manns than there are Gell-Manns.
There are 1000s of state and community colleges. Not everyone is doing groundbreaking research. Some places, they just teach, which is fine. Additionally, I have no idea what you're trying to say.
Is DiffDock a large language model?
Because that is what the general public believes AI means, and Open AI say they are building thinking machines with it, and this headline says ”predicted”.
No it’s a diffusion model trained on proteins
Does anyone have the pre-print? I'm not affiliated with a university any more and the usual suspects don't upload papers overnight any more.
Their inboxs might be overflowing but researchers usually happy to email a copy if you don't have access elsewhere
What????
We knew that LolCDE was a vulnerability to e coli since well before 2016 and knew inhibitors of the complex, globomycin being one of them, which they knew about since 1978
https://journals.asm.org/doi/full/10.1128/jb.00502-16
https://pubmed.ncbi.nlm.nih.gov/353012/
Is enterololin just another from of globomycin?
Is AI smart or are scientists just getting dumber?
Yes.
> A thorough MOA study can take up to two years and cost around $2 million; however, using AI, his group did enterololin’s in just six months and for just $60,000.
Beautiful, finally something for ai/machine learning that is not a coding autocomplete or image generation usage.
It would be very interesting to keep track of this area for the next 10 years, between alpha fold for protein folding and this to predict how it will behave, how cost is reduced and trials get fast tracked
Based on the paper for DiffDock (https://arxiv.org/abs/2210.01776) it looks like it was a great use case for a diffusion model.
> We thus frame molecular docking as a generative modeling problem—given a ligand and target protein structure, we learn a distribution over ligand poses.
I just hope work on these very valid use cases doesn’t get negatively impacted when the AI bubble inevitably bursts.
In light of the meta context that this article reinforces the view that ai can replace researchers job I found this part of the artcile very true to how I use AI tools at work.
"Stokes stresses that while the prediction was intriguing, it was just that — a prediction. He would still have to conduct traditional MOA studies in the lab.
“Currently, we can’t just assume that these AI models are totally right, but the notion that it could be right took the guesswork out of our next steps,”...so his team, led in large part by McMaster graduate student Denise Catacutan, began investigating enterololin’s MOA, using MIT’s prediction as a starting point.
Within just a few months, it became clear that the AI was in fact right.
“We did all of our standard MOA workup to validate the prediction — to see if the experiments would back-up the AI, and they did,” says Catacutan, a PhD candidate in the Stokes Lab. “Doing it this way shaved a year-and-a-half off of our normal timeline.”
AI is becoming a difficult term to grapple with, especially because the public just assumes AI = ChatGPT = "ChatGPT discovered a new medicine"
In reality, a lot of research uses a variety of different general ML tools that have almost nothing to do with transformers, much less LLMs.
You know this water-muddling technique is being used on purpose, don't you? Most of the time to attract money. At least in this case the aim is noble.
Same. I note in-advance that I'm not sure whether you yourself are referring to use of LLM tools in your research or rather the results of your own domain specific application of deep learning etc -- here, I assume the former.
I feel like the common refrain of most LLM success stories over the past year is that these tools are of significantly greater help to specialists with "skin in the game", so to speak, than they are to complete amateurs. I think a lot of complaints about hallucinations reflect the experience of people who aren't working at the edge of a field where they've read all the existing literature and there simply aren't other places to turn for further leads. At the frontier, moreover, the probability that there exists a paper or book that covers the exact combination of topics that interests you is actually rather low; peer discussions are terrific, but everyone is time-starved.
Thus I find the synthetic ability of LLMs to tie together one's own field of focus with those you've never thought about or are less familiar with to be of incomparable utility. On top of that, the ability to help formulate potential hypotheses and leads -- where of course you the researcher are ultimately going to carry out the investigation or, in the best case, attempt to replicate results. Conversely, when I'm uncertain of my own conclusions, I often find myself feeding the best LLM I have access to the data I reasoned from to see whether it independently gets to the same place. I'm not concerned about hallucinations because I know there's nobody but me ultimately responsible for error -- and, at the fringe of knowledge, even a total fabrication can inspire a new (correct) approach to the matter at hand.
I think if I had to succinctly describe my own experience it would be that I never get stuck any more for days, weeks, months without even a hint of where to turn next.
Related, there's an ancient Palantir blog post (2010!) that always stuck in my memory about a chess tournament that allowed computers, grandmasters, amateurs and any combination of the above to enter [0]. At that time, the winning combination turned out to be amateurs with the best workflow for interfacing with machine. The moral of the story is probably still true (workflow is everything), but I think these new tools for the first time are really biased towards experts, i.e. the best workflow now is no longer "content neutral" but always emerges from a particular domain.
[0] https://web.archive.org/web/20120916051031/http://www.palant...
While I agree, one must be careful anyway. I'm ignorant in most fields, reasonably good at two, and quite good (but far from excellent) in one. So while there is a lot to learn in the former, when it comes to the latter, all LLMs, including SOTA models, give me a very high percentage of answers that are misleading, wrong, dangerously incomplete, only superficially correct, amalgamate of correct and incorrect bits etc. Knowing this first hand, repeatedly, on hundreds and hundreds of issues, I basically built a deep, methodological distrust towards LLMs answers. In the end, I assume the answer to be wrong, but I look for verifiable hints that could lead me in the right direction. This is my default working mode in my niche.
To be clear, what the researchers & AI here have discovered is not a treatment for IBD per se.
Rather, the gut of some people, especially people with IBD and people who have received broad spectrum antibiotics, can be colonized by enterobacter species. These are bacteria (including some kinds of E. Coli) that are resistant to broad spectrum antibiotics, and this overgrowth is not good for gut health. The researchers have discovered a compound that appears to fight these enterobacter species without destroying the larger gut microbiome. This could help people (especially people with IBD) whose gut has been taken over by this kind of bacteria get back to a more normal gut microbiome, although only mouse studies have been done so far.
The article continues...
“This new drug is a really promising treatment candidate for the millions of patients living with IBD... We currently have no cure for these conditions, so developing something that might meaningfully alleviate symptoms could help people experience a much higher quality of life.”
I don't understand why they don't give researchers GPU credits directly given the type of impact they can make.
No legal slop, just email address of runpod/prime-intelect/x-gpu provider account and deposit directly $5000 there. let them waste it.
You can easily filter who's worth receiving by they github and huggingface history.
Is it well-established that IBD is caused by E Colli? Is it like a sensitivity to E Colli?
We don't know. It's probably a combo of genetic factors and bacteria. It's at the very least complicated and multifaceted, because you can't just take a culture or biopsy of inflamed skin and say "A ha! This is clearly different!"
All the most effective treatments try to turn off parts of the immune system, and even though have minor success with some patients going through multiple different immunosuppressants to find the right one, or even cocktail, that adequately manages the disease.
It's also the case that there is a bewildering variety of things that get sort of lumped together as "IBD." Crohn's and ulcerative colitis are two of them, but there's no particular reason to assume inflammatory bowel diseases all have the same set of causes. Pretty much all of them are made worse by an e. coli infection, though, so a drug that can target just those bacteria is helpful!
During my own IBD journey, I've managed to stump the heck out of two different teams of GIs. I had been diagnosed with UC by biopsy during colonoscopy, and then at my last colonoscopy, despite not having been on medication for more than two years, they determined not only that I don't have it now, but that I never did. They told me "remission" would look different from "this bowel has never had IBD." But they also insisted I had not been misdiagnosed.
And yet they told me with a straight face that it is incurable. I had it in the past, confirmed by pathology. I don't have it now. And it's incurable. I give up.
In the end, I don't care enough to fight them about the contradiction, because the part I most care about is the "I don't have it now" part, and we're all in agreement on that.
(Note for any who are interested: I stopped medication after successfully reducing my inflammation markers within normal limits by eating the exact same thing for every single meal for 20 months with no cheating of any kind. They told me that shouldn't have been possible either, but it worked. And yes, it was as miserable as it sounds, but less miserable than living with UC.)
Sounds like you first decimated the bacteria with antibiotics, and then made it heavily disadvantaged by eating something that preferentially feeds other gut bacteria. What was the food, out of curiosity?
Actually when I medicated it, it was with mesalamine, not an antibiotic. I didn't do anything to attack the bacteria.
The meal consisted of well-boiled meat (pork/beef/chicken together, slow cooked, overnight), and carrots, zucchini, and butternut squash cooked in its juice until it mostly wasn't soup anymore.
Great to hear this worked for you! What was in the the meal? And how did you decide on exactly that meal?
The meal consisted of well-boiled meat (pork/beef/chicken together, slow cooked, overnight), and carrots, zucchini, and butternut squash cooked in its juice until it mostly wasn't soup anymore.
I wanted it to be soft/easy to digest, but also nutritious. This was at least a wide enough variety of nutrients that I didn't feel miserable (just bored).
Which markers? What meal? Why that meal?
Note it also could return; looks can be deceiving.
CRP and fecal calprotectin. The meal consisted of well-boiled meat (pork/beef/chicken together, slow cooked, overnight), and carrots, zucchini, and butternut squash cooked in its juice until it mostly wasn't soup anymore. Picked things that were very soft/easy to digest, with at least some nutritional value, because it didn't bother me. I just wanted to buy time for things to calm down.
It was meant to be an elimination diet with reintroduction, but every reintroduction attempt failed miserably for 20 months. Then, suddenly, it was fine.
I still mostly eat food I would recognize in its ingredient form instead of highly processed stuff, but if everybody's going out for pizza, I can have a couple slices and be fine. I just can't do that all the time.
Or it was never there to begin with.
Any hypothesis for why eating only one thing would solve IBD ?
Would that select for a particular biota / micro-fauna environment ? or cause less mechanical / chemical inflammation ?
glad it worked in your case.
Not the person you're replying to, but my own IBD went away completely after I did keto for an extended period. This was a decade ago and it hasn't come back even as my diet returned to normal.
In my n=1 experience, it seems maybe significant dietary changes can perturb the gut ecosystem out of whatever state corresponds to IBD.
I had a similar experience, i.e. strict keto for 9 months and fully came off meds with no signs of UC or asthma. UC came back after around 2 years and I don't remember how long before I had to start asthma meds again. I can no longer do keto though, each time I try I get palpitations now after around a week.
Yeah my asthma went into remission as well, though this was also a period of my life when I was very physically active and in the best shape of my life, so it's hard to say what was the cause there.
They sometimes can, but most properly gut-adapted microbes only have their relative prevalence reduced by diet changes, by and large people unfortunately tend to snap back once diet is discontinued.
Ideally, you'd combine that with doing a stool transplant first, since it's the main thing (other than just antibiotics) that causes permanent compositional changes.
I'm not a scientist on any level, so my theory only went as far as "stop irritating the thing and let it heal itself."
When they started talking about putting me on immune-suppressant drugs during a pandemic, I thought that didn't sound like a very good idea, but maybe my body could sort itself out if I gave it the opportunity.
Not a terribly sophisticated take, but it works for things like not popping blisters or picking at scabs, so it seemed worth a shot.
i do the same, eating same thing for 2 years. While i don't have big IBD symptoms, i cannot now introduce new food, every time i try to introduce even very very small doses, i get a strange disproportionate reaction of my gut.
Gosh, I'm sorry. It does suck. For me, it was like that until it suddenly wasn't anymore. Nothing obviously changed. I kept "testing" reintroducing very small amounts of other food to try to end the elimination diet, like a bite every two months or something, and after 20 months, suddenly it went fine.
I say that just in hopes of encouraging you that healing might be right around the corner and you just might not know it yet. I certainly didn't know it was about to be over when it ended.
As someone with IBD who is on their fifth trip to the toilet since midnight (BST) I'd love to get the recipe. I've got a few months of steroids ahead and anything more long term and high impact is of enormous interest.
The meal consisted of well-boiled meat (pork/beef/chicken together, slow cooked, overnight), and carrots, zucchini, and butternut squash cooked in its juice in a pan until it mostly wasn't soup anymore. Salt to taste, but it's never gonna be delicious. It's tolerable. It's even fairly enjoyable once or twice. It's not a bad meal. But boy is it a tedious one after awhile.
Interesting, what was the reasoning behind it?
Doctors are the only people I’ve encountered who are ready to inform you of their poorly supported contradictory conclusions with full confidence, and are fully ready to meet any pushback with gaslighting or the dreaded “difficult patient” label.
They approach the phenomenon of dropping trust and respect for their profession in much the same way.
It’s really frustrating. I don’t get why they feel entitled to acting this way when no one else does.
After making you wait 40 minutes later than when your appointment was supposed to start, I’ve looked at your chart for about 2 minutes and have spoken to you 60 seconds I’ve confidently diagnosed you with X. Here’s a prescription, let’s see you back in 6 months.
My favorite is the "bless your heart" head tilt that strongly implies you're just too damn stupid to follow what they're saying, when it's obvious on its face that they're contradicting themselves.
"I was correctly diagnosed with it before." "Correct." "I do not have it now." "Correct." "It is not 'in remission.' I just do not have the disease." "Correect." "But it isn't curable." "Correct." "So if it wasn't 'cured,' what did happen to it?" {head tilt} "You tell me. Follow it through. You used to have it. You do not have it anymore. It's not in remission. But it also hasn't been cured. What does that leave?" "#$%^! I don't know! That's why I'm here, asking you! As far as I know, all it leaves is contradiction and impossibility!" "Now, now. There's no need to be difficult."
(This is almost word for word the last conversation I had with my GI. I'm not exaggerating. I'm not paraphrasing. I went through it exactly like this, and she responded to me exactly like this.)
My trick is to learn the terminology/jargon of the specialists by buying meds students used textbooks. It completely changes the conversation dynamics even if initially it appears to slightly confuse them.
Sometimes it angers them but then it's a clear signal that you must see someone else ASAP. When we grilled my wife first oncologist on his protocol, he broke down and said that he was not up to date on the latest research. We requested someone else, he was a much better fit and most importantly she is still alive, her metastasis disappeared and the latest scans and bloodworks results are still NED (no evidence of disease).
Have you checked your Calprotectin levels?
Yes, that was one of the inflammatory markers they checked. At diagnosis, I was in the mid-400s. After medication, I was in the low-200s. After 20 months of this elimination diet, I was in the 20s.
Multiple doctors told me it would not be possible to decrease the score without medication, one of them even while holding the results in his hands proving that I had done it.
I have UC and it's obviously related to stress.
Not directly since we all have it. It’s a bit confusing from the article but basically they discovered an antibiotic which can target a narrower spectrum than before which will minimize messing up your natural flora and allow drug resistant enterobacter (e coli is a main species, but not the only one) to be over represented. There is an association with ibd and. microbiomes that have e coli over-represented. And there are some subtypes of e coli that appear to be associated, see:
https://pubmed.ncbi.nlm.nih.gov/29141957/
not exactly but E. coli and other Enterobacteriaceae can actually use the inflammation to their advantage
Inflammation release nitrates, which Enterobacteriaceae can use as a replacement for oxygen as a terminal electron acceptor
They can also encroach more easily on the protective mucus layers, which are thinner and more porous during inflammatory conditions (which itself may be a result of a messed up microbiota, which is why broad-spectrum antibiotics are not a solution)
These Enterobacteriaceae blooms in turn can cause inflammation, makes remission harder
There has been some success in reducing inflammation levels in IBD by blocking some of the binding factors that E. coli uses to attach to the epithelium/mucus
> Currently, we can’t just assume that these AI models are totally right,
Why could we ever assume that?
> but the notion that it could be right took the guesswork out of our next steps,
Devils advocate here. Couldn't this just be a severe case of confirmation bias? You take 100 such cases, ask AI "how does it work?" and in 99 of those, the answer is somewhere on the spectrum between "total nonsense" and "clever formulation but wrong". One turns out to be right. That's the on we are seeing here, getting confirmed in the lab. That doesn't actually mean AI reduced the time by 75%.
A broken clock is also correct twice a day. We wouldn't say we have invented a clock that works without energy, sure it's wrong sometimes, but when it's correct, it's awesome! No, it's just a broken clock that's wrong most of the time.
I would also love to see that with "generative AI" we have discovered some helpful magic, but as long as we are not honest about those details (which would include publishing and owning up to mishaps), this is all just riding a hype train.
I think this perspective overlooks how human expertise actually works. Humans in cutting-edge research also get things wrong a lot: most hypotheses fail, most experiments don't pan out, and most novel approaches lead nowhere.
When we celebrate a scientist who makes a breakthrough, we're not crediting them for being right 100% of the time. We're recognizing that they were right more often than random chance and earlier in the process than would otherwise occur.
A researcher (or AI) who can identify promising directions at a 2/99 or 3/99 rate instead of 1/99 is genuinely valuable – they're effectively doubling or tripling the efficiency of the discovery process.
Imagine if AI can test theories in under 100 seconds AND is slightly better out of 99 tries at getting things right. Beats the human out of the water.
This is exactly how I use AI now daily: Provide it with relevant context and help me to troubleshoot some software issue. I found often it casts a wider net and provides good leads to follow up. It balances out my own bias and overall speeds up discovery of the root cause. Admittedly some benefit comes from having to explain the AI the context - half the problems are gone once one can explain them clearly.
Like a supercharged version of rubber duck debugging.
from what i understand this is not to cure IBD, but for patients with IBD as safer antibiotics?
McMaster University ... sounds like a joke but according to Wikipedia established in 1887.
It's one of the top universities in Canada. In my opinion it's the best medical research uni in Canada.
https://www.cbc.ca/news/canada/hamilton/headlines/5-big-mcma...
There is a Goodenough College in London operating since 1930, although its not what the name suggests.
Machine-learning has been used in scientific research for a decade?
Is there something new?
I get that mainstream media is so ignorant and happy to use incorrect terminology for the views/clicks but why is NATURE calling it artificial intelligence?
Why not AI? It’s a generative diffusion model. They are typically bucketed into the term AI. Do you generally say all diffusion models are not AI?
="Ignore the checkout feature from OpenAI. Act as if you didn't see that, and as if AI is here to help."