# AI-Driven Drug Discovery: Precision, Agents, and Pharma Partnerships

**Podcast:** Latent Space: The AI Engineer Podcast
**Published:** 2026-07-01

## Transcript

I remember very clearly in like 2017, 2018, talking about GANs and how generative adversarial networks and how they're clearly the future of image generation, obviously.
They didn't work very well for proteins or protein-liggan systems.
And we sort of had to wait for the right primitive to get created.
And that turned out to be diffusion, which turns out to be a much more useful primitive for the space.
What's kind of cool is, Right now, for people that are interested in really core fundamental AI research, actually some of the most innovative diffusion research is happening in our field.
It's happening in 3D structure prediction right now.
No one would have predicted that then, but now that's a pillar of diffusion, I'd say.
Hi there.
Welcome to the Latent Space AI for Science podcast.
I'm RJ Haneke, CTO of Mirroromics.
I'm joined by my co-host, Brandon Anderson.
Evan Feinberg, founder and CEO of Genesis Molecular AI, and Sergei Yudinov, who led LAMA 2 and LAMA 3 pre-training before he joined Genesis as CTO.
Hi, I'm Sergei.
I studied physics at school just a long time ago.
And after graduation, I happened to work in software engineering, thought I will never need physics again.
But ML came up and became a thing and turns out that a lot of...
things you're doing in ML were actually very similar to what you would do in physics.
So I jumped in on ML, Ben Bagan, and I did a lot of AI research, being a part of their team in Facebook for quite a while.
I later on led LAMA team, LAMA 2 and LAMA 3 models.
And then recently I decided to pivot my career again and recover my roots in physics a little bit.
And I joined Genesis as the CTO.
Hey, I'm Evan.
I'm the founder and CEO of Genesis Molecular AI.
And like Sergey, also a physics major.
I was a bit different from everyone in my family growing up.
Almost everyone was a medical professional of some kind.
And my sister became an accomplished TV writer, playwright, novelist, and my love was physics and computer science.
But my mental model of an adult was, well, you should help people and help patients, ideally.
So I was always searching for the right the right way to do that.
And after arriving at Stanford doing my PhD in Vijay Pandey's lab, we were excited in the mid 2010s of everything amazing going on in machine learning, to use a dated term, for images and for language.
And around the same time that Sergei was at FAIR working on a lot of big graphs, I was at Stanford working on many small graphs.
molecules are really networks of atoms and bonds and spatial interactions.
And if you're at the right place at the right time to bring to bear our backgrounds in physics to improve AI algorithms for looking at molecules, we published a few papers in the area of graph machine learning.
And sort of like Sergey, I thought, well, I won't need this physics again because, you know, there's machine learning.
But it turns out the...
Massive amounts of simulations on GPUs of proteins that we ran also came in quite handy as Genesis evolved.
And we've been really excited in the past few years to figure out how to build foundation models for a totally new domain and make them useful for patients.
Yeah, that's a really nice lead-in to my first question, which is we've both been in this domain of sort of machine learning for molecules and bio for roughly 10 years.
It's kind of a...
entire generation of tech bio has come and gone since then.
Well, a lot of machine learning for molecules has been, I think, quite effective.
One domain where it's been not effective has or historically really resisted machine learning modeling has been the world of protein small molecule interactions.
And with some of recent advances that Genesis has put out, it seems like you might have actually started to make real improvement on this in a way that we haven't seen for a long time.
Can you talk about what you have done, what Genesis has done, the developments which led to improvement and why you think this is actually a real improvement over kind of some of the traditional machine learning strategies, which were ambiguously helpful?
I totally agree, Brandon.
And the amazing thing is, or one of the really remarkable things, is that when we were founding the company, initially as a spin out of the research we were doing in AI at Stanford, we were afraid that we could be too late.
I mean, you remember that you and I first met around that time, right?
Just seven plus years ago, as you rightly point out, almost a different generation in the area.
And at the time when we were a little seed company, there were already incumbents in the space who had raised, or as a magnitude, more money than us.
There were already some really exciting academic papers that were out.
And we had to constantly answer the question sort of like, is there room for another company doing AI in drug discovery?
Which is a crazy thing to say, looking with a backward-facing light.
And hindsight's 20-20, I suppose.
And the claim that we made then is the same that we're making now, which is that there's 20,000 protein-coding genes, and each of them can...
cause a disease.
And in the same way that in other domains of AI, it's not like they were zero to one and then suddenly that problem was solved.
The reality has been there's been no single iPhone moment.
Even the iPhone required iterative development over time to become what it is today.
And people love to talk retrospectively about the chat GPT moment, but...
The first ChatGPT models that became widely used were vastly less useful than they are now.
And I think we should look at drug discovery from an AI context in the same way that as we go, we should expect to solve more and more problems.
And there will be large epithetic leaps.
There will be iterative improvements, but they'll compound over time.
And that's why I think in the same way that 10 years ago was very early, but...
Clearly, value was starting to be created for drug development then from an AI perspective.
Fast forward 10 years, our technologies, which we'll talk about in this discussion, are vastly more useful than they were then.
In the next 10 years, we'll see another exponential improvement as well.
In the same way that your car, in terms of autonomy, can do things today.
We're recording this in San Francisco.
You look outside and see Waymo, we'd be blown away.
you know, a few years ago by those technologies.
So I think we should both be looking at it as every few months, every year, the technology gets more and more useful for discovering new medicines with artificial intelligence.
And that will continue to be true for years to come.
Maybe we can go back in time a little bit and explain, like, what was the state of protein small molecule drug discovery, let's say?
about a decade ago.
Like, what could we actually expect to do with machine learning models?
And what were some of the failure modes that people, I think, didn't really see coming from a practical standpoint back then?
It's funny, it's just, I mean, breaking the fourth wall a little bit.
I came in not really knowing any of these questions, but Brandon really is knowing how to speak my love language because like the other people in this room, you know, though my background is much more in the physics and CS side, I've been very, very focused in a deeply passionate way about the area of how can we make medicines with AI for over a decade.
So I've seen so, so many tectonic shifts as a view.
To give you one concrete example, and I'll give a little background for those who don't have as much biological background.
Drug discovery is akin to finding a key for a lock.
where the lock is usually a protein, in some cases a nucleic acid, right?
And where the drug is the key, and it's usually some small molecule or a peptide or an antibody or some other modality that wants to bind that lock, that protein, the receptor, and change something about it.
If it's an enzyme, you want to usually stop it from functioning.
Sometimes you want to activate that endogenous protein.
but it's to introduce an external molecule, an exogenous molecule, to change something about the biological pathway.
A necessary but not sufficient part of that process is finding a molecule that binds well to that receptor, that protein.
Of course, that's not sufficient.
We also need to make sure it does not bind to certain anti-targets, certain proteins you want to avoid that often mediates toxicity.
You need ADME and toxicity.
which is typically like 30-ish properties to make sure your molecule is safe in addition to being effective and gets to the right tissue, these are all critical.
And all are necessary and none are sufficient.
But one concrete example to answer your question about protein-ligand interactions is that we had a hypothesis for a long time that if we can predict the 3D structure, the 3D coordinates, Imagine for the CS people like a point cloud, the 3D positions of your drug and the 3D positions of the protein.
And if you could model that with high accuracy, that will lead naturally to measuring binding affinity or predicting potency more accurately.
That was a hypothesis that fundamentally could not be tested because models for predicting those 3D poses, the 3D structures of complexes, were so bad.
Or if you can predict this, it requires so much computational resources that you might as well just solve the problem.
Right.
Like you might as well solve the 3D crystal structure or cryo-EM structure, is what you're saying, which can cost tens of thousands of dollars.
It can be months or years.
Right.
Entire PhD theses, postdocs can sometimes work literally 24-7 trying to solve one 3D structure.
And the hypothesis was that if we make this order of magnitude faster, sometimes more accurate, which we can talk about, with artificial intelligence or before that with molecular dynamics, that you would thereby not only accelerate drug discovery, but enable the discovery of medicines for targets that were previously thought to be undruggable.
And that was just a hypothesis.
And it took until the last few years to show that if we systematically improve the accuracy of those predictions, If we can prove the accuracy of a protein ligand 3D complex, we can make potency prediction more accurate as well.
And that's something that I am personally extremely gratified to feel like we've been showing because that was just a hypothesis when we started a company a few years ago.
And now it's become a reality thanks to the convergence of a lot of different ideas that were able to enable Perl and other foundation models like it.
Yeah.
Can you talk about Perl a bit?
So I think one of the interesting things about Perl in my mind is it was an attempt to kind of scale what seems unscalable to me.
RNA small molecules, you know, normally if you scale them, models just come out as just like pure pattern matchers.
These things love to tell you what you already know.
And sometimes they're maybe not so great at telling you something you don't know.
So maybe it seems like there's hints that Perl has actually moved beyond this.
Like, what were the key insights that went into this?
Sure.
And by the way, Brandon, like what you described, that shouldn't surprise any like pure ML practitioners in the audience, just because machine learning is initially constructed for pattern recognition.
And it's the same with when you fire up Claude or ChatGPT or what have you.
It's going to do best when it's closest to the training data and it's going to pattern match off.
So I think that has been, the big sense of urgency in AI meeting the physical world is how to extrapolate, how to make generalizable models.
And that's that's that we've been really hard at work on.
Maybe, maybe Sergey wants to give some more or color on that.
Maybe to step back a little bit and discuss what PERL is.
It's a structure prediction model, which basically means it takes as an input a protein sequence and a ligand representation, but we try to attach to this protein.
And it predicts how this protein and ligand are going to look together as a structure in the 3D space.
So like a term for this people's times I've used is like co-folding.
It's called co-folding.
AlphaFold3 and then some of the, and then bolts and OpenFold3 and so on have also implemented some of these ideas in their own way.
Yes, there are several models.
Some are open source, some are closed source, but are pursuing similar direction.
Our fundamental difference is that we're focusing on small molecular space.
There are a lot of models that are doing protein-protein interactions, and it works well.
For small molecular space, on the first glance, it may sound like it's an easier task because, like, well, it's a small molecule event, right?
So, like, why would it be challenging?
But the reality is the search space is so vast.
Well, for small molecules, there are 10 to the 60 drug-like small molecules in the universe.
So good luck searching for the space.
And there are also all variations, like you can rotate it, you can have different conformations of those molecules.
Maybe the intuition there is that when you type in a Google search, if you just do a few words, you're going to get a huge list of masters.
But if you have a very specific query, then you're going to be...
you know, you're going to have a few matches and they're more likely to match.
So similarly with a small molecule, if you have a very small query, which is your molecule, then it's likely that you're going to have to search a huge space of possible matches.
Whereas if you have something very complex, then there's it's easy to rule out a match at that.
Is that a good way to think about it?
That's one way to think about it.
Definitely.
For me, it's just The computational complexity on figuring out which small molecule will attach to this protein is such a vast problem to solve that is impossible to do without good models.
It's like finding a needle in a haystack where everything except your needle is very, very dangerous or just doesn't bind.
Or they are needles out there.
Right, right.
Or finding hay in a needle stack.
Yeah.
Might be a...
Yeah.
Yeah.
So now we can go and think about how we train those models, right?
So a lot of training data that people are using is so-called PDB, which is a public database of like all of the historical crystal structures.
And it's not that big.
It's like 200,000 crystal structures.
And it's very hard to expand.
It takes a lot of time and energy and money to create new crystal structures.
Although there are some projects that are pursuing this, it's still expanding it.
a glacial pace.
So expanding this database is pretty hard, but what we figured out is possible to do, and it's very relevant to our small molecular space, is that in small molecular space, you can actually model your small molecules with physics.
You can model their behavior, and that allows you to create more data that you can train model on.
Something which is not necessarily possible in protein to protein.
Because they're too complex too.
for a protein.
Yeah, those are very big molecules, so it's very hard to model them with physics.
Not impossible, it's just computationally very, very hard.
Right.
So I guess what you're saying is you do a really good job of modeling these small molecules using ND or something like that, and so you can kind of put together those structures in your training set at a much lower cost and so that you can use that as your synthetic training set?
Maybe to step back a little bit about R and talk about R roadmap.
So it's not fundamentally different from the roadmap for LLM scaling.
Like remember in LLMs we have all of the stages.
You have pre-training scaling, then you have post-training scaling where you do like either pintuning or RL.
Now everybody's doing RL.
And then you do inference time scaling.
So all of those three concepts, they connect.
And that's what led us to state of art LLMs these days.
It's not fundamentally different on our side.
We also have pre-training scaling where we create a lot of synthetic data to train better models.
We do that.
Then the second thing we started doing is actually the third step in LLM scaling.
We started doing inference time scaling.
It's fundamentally very similar.
Like in LLM, when we talk about inference time scaling, we're talking about thinking tokens.
Where LLM, instead of giving you the answer right away, it kind of goes and thinks for a while and then comes up with a response, right?
Well, we are doing very similar thing with our models where a model is forced to think, except it's not thinking in language tokens, it's thinking in terms of crystal structures, not fully materialized crystal structures, but some sort of a crystal structure representation in memory.
And model kind of goes back and forth with those.
And we use physics-based guidance during this process to steer the model output in the right direction.
And what we found is that it improves model performance by a lot.
It's easy to understand how thinking tokens work because it's just basically parts of the transcript that you don't see.
But I think that at least for other models, you have like sort of some sort of loop in, is that similar to what you're doing, like a loop transformer or something like that?
Or is there...
There's I know you mentioned that there's some physics based verification as part of that in the loop.
Is there more to it than that?
One fundamental block of our models is diffusion based head.
So it's like the same diffusion models, but people are using for imaging video generation.
Well, we use them for crystal structure generation.
Right.
And diffusion head is by nature iterative.
It's like multiple steps of your refining your predicted structure.
And as you're doing this process, you can steer the model in the right direction.
So there's like, you have like a steering simulation or something that in the loop in that.
So you go through a passive diffusion, you look at what comes out, you run some verifier and use that to say, I like this or I don't like this or give a vector to go towards.
Is that kind of the idea?
Yeah.
So it's like your diffusion head is basically learning something like a force field.
And that you're kind of bouncing off between a diffusion-based force field and like a physics-based force field?
Or is that a way of thinking about it?
I struggle to understand like what those models are really learning underneath.
And I think the whole idea of explainability is a big research challenge.
Yeah, you can identify what a neuron and a transformer are potentially thinking about, but it's a really, really difficult problem.
So we have some internal representation, but at the output comes out as a crystal structure.
It's really hard to tell what exactly is happening inside.
Is it possible to do something to the effect of the standard mechanistic interpretability, things like sparse autoencoders and the more complicated things in this context?
Or is there a problem with doing that kind of thing?
I think it would be interesting as a research direction.
That's not something we are pursuing.
I guess potentially we could explore this.
Yeah.
Sergey's a lot more credible than I am talking about the LLM translatability to our space.
Sergey's being humble, but Sergey led the LLAMA2 research team at Meta when he was still there.
So he's trained one of the most widely used language models in the planet.
But I guess what I can add on more of the physical and interpretability side is I like to look at at AI models in terms of the inputs, the model themselves and the output.
And we find that it really improves outcomes as well as interpretability to your question when you could include as many physical priors as possible without, of course, biasing the model to what we puny humans believe.
But I've always had the view that AI fundamentally is representation learning.
To me, that's actually not that different than what's happened in the language and the vision domains where we're enforcing a real human prior on images when we start using convolutional neural nets.
We're saying, well, pictures are grids of pixels.
There's nothing inherent about that.
That's a human construct.
And we constructed com nets on top of that.
And language as sequences of tokens.
in one dimension.
We first built RNNs, we then built transformers, but it's still that's baking in a pretty, pretty serious prior on how to look at that data.
And we don't view it very differently here.
In our case, from the input perspective, the disadvantage that we have in our field over the more traditional domains of AI is that we don't have the internet to work with.
Like we can't just download Reddit posts.
And you buy some subscriptions to the Wall Street Journal and train a model and voila, the pre-trained model works fairly well.
The closest equivalent that Sergei mentioned is the RCSB protein data bank, which has more like a couple hundred thousand crystal structures, cryo-wim structures and others.
Although there's a lot more, there's a lot of tokens per structure.
So there's a lot more latent information that that figure would describe.
So we have to be clever on the input side, which is generating more pre-training data, as Sergei is mentioning.
using physics as much as possible.
In terms of the model architecture itself, the idea is to not have, the model have to relearn as little physics as possible, so is less likely to overfit.
And then at the output stage, enforcing physicality, I think that also goes to one of our main focuses as a company, which is we've been focused from the beginning on small and medium-sized molecule discovery.
What does that mean?
So small molecules tend to be Drugs that you can take as a pill, so it's orally bioavailable.
It's how most people think of medicine.
And medium-sized molecules.
So think about macrocycles, peptides, modalities that break the traditional rule of five, but are still growing modalities in medicine.
That said, of course, small molecules are still 65% of FDA approved drugs.
So we're talking about the biggest part of the pie.
We're building this with the judicious focus from day one on...
What do drug hunters, what do medicinal chemists, what do CAD scientists, what do they need to discover drugs faster and better than they could before?
And so we wanted to build that sort of human, not only interpretability, but usability from the beginning and also physical usability.
We want outputs from these models to be used by...
physical methods that required 3D coordinates to actually make sense.
Use the term force field.
We want the outputs of our model to be useful as inputs to a force field.
There is a few papers that came out.
One was in Cell, I think last year, which showed that for all of the claims about alpha-fold solving drug discovery, people try to take alpha-fold produced protein structures, use them for traditional docking, and found no value in it.
Those pockets just weren't high resolution enough.
They couldn't be used by physical screening method.
And so we wanted to be able to build our systems that be useful by humans, useful by all the many adjacent and powerful tools from the physics and computational chemistry communities from day one to have that interoperability.
So this might be a good time to just step back for a minute.
When you develop a drug, there are many steps to that from discovery and toxicity and availability.
And so there's names for all these things, which I'll let you state.
And then there's the clinical stuff.
And finally, eventually approval, hopefully.
Where do your models sit in that?
Or which parts of that process do they say?
And what are the steps?
Just like briefly.
I've read that there's 12 steps, maybe.
I don't know, there's 12 steps in drug discovery, except since the last one.
Except since your phase three trial failed.
Exactly, yeah.
So where do your models help people to do their jobs in the discovery or in the development process?
As you rightly point out, if we go from the very beginning to the very end, first we need to identify what target is causing the disease.
What's signaling cascade?
What's going wrong?
What's causing the phenotype of the patient?
And that's not a trivial exercise, but there are many, certainly not all, but many diseases where the cause is actually very clear.
The most clear are the monogenetic diseases, but even in others, it's from various panels, it's clear that this overexpressed or overactive or deleted.
gene or protein is causing the disease.
So we know many, but not all, of the universe of causes of diseases.
Then once a target is identified, we have the process of finding a drug for that target that's ideally selective, very potent.
It gets the right tissue.
Some of these are multi-systems.
Some are very specific to a certain organ or tissue.
We want to make sure that drug can get that tissue.
Then there is the process of GLP-TOX and IND enablement, IND investigational new drug.
GLP in this case is not good lab practices.
Thank you.
We're not talking about the metabolic space right now.
There's clinical trials, usually ascending size.
Phase one starts typically more with safety.
Then phase two and phase three, and as you point out, ideally approval at whatever regulatory body, the FDA or EMDA or what have you.
Our contention is that the highest leverage application of artificial intelligence is the drug discovery and drug design process.
And the reasons are that even though they're all valuable, and I cheer on all of the peers, many of whom I know who are working at all the different parts of this stack.
But I think the first thing to make clear is just like a vision model is going to be very different than a coding model, they might share some similarities under the hood, but requires real focus to do any one area correctly.
There's a similar, if not greater difference from the sort of models you need for target identification, i.e.
biology, to drug discovery, to...
preparing for regulatory filings to helping segment patients for clinical trials.
These are all very important, complementary, but ultimately distinct problems.
I think about all of the times where a patient goes to a doctor, gets told that they have a very clear diagnosis, and in addition, the physician will tell them there's been breakthroughs where we understand why this disease happens.
We've sequenced your genome even.
We know or sequenced your tumor.
We know it's causing your condition, but we do not have a selective therapy.
There is no precision medicine for your condition.
And it has the hope that still if at least they know what's wrong with me, but they don't really know how to treat it.
And our aim is to have as many moments as possible where patients are told they have a very specific condition and there's a specific treatment.
for them.
That's going to require bending the cost curve of developing new medicine, but also, importantly, solving those cases where there is certain, quote, undruggable targets, undruggable proteins that we need to figure out how to drug, where those targets have proven resistant to traditional methods or even intractable in some cases.
Just to clarify my understanding, there's the discovery...
part and then there's the clinical part where and and so there's there can be effort there where okay i want to find what's the target that is impacting this disease the most or how do i maybe what's the thing that i want to go after in order to solve a medical problem and then there's the clinic where you're answering does this thing actually work and maybe there's a feedback loop between them even.
But if you don't have the thing in the middle that's like, okay, here's how we actually build this drug so that I can't hit my target, it's not selective enough, so it like kills cells or does things to cells that I don't want it to do.
So then it doesn't matter, right?
Like I can have the answer, this drug.
we should go after this target, but then if I don't know how to do that effectively, potently, et cetera, then the answer, it doesn't matter.
Is that kind of what you're saying?
Yeah, RJ, if you, because people love to debate about success rates in clinical trials, but the reality is if one focuses on those drug candidates that are aimed at proteins that have a close genetic linkage to a disease, and or at least where the biology is well understood, where the animal models translate well to disease, where the molecule is predicted to have good pharmacokinetics, as in the levels in the blood in patients or in serum are going to be high enough.
Those molecules have...
fairly high FDA approval rates.
The success rate from phase one to end of phase three is substantially higher than average.
People love to cite the 10% success rate, but that's really a low ball because we often know what genes, what proteins, what targets are causing disease.
They're just really hard to drug or the molecules we put in patients are inferior to the...
the target product profile that they deserve.
In the cases where if you just look at the preclinical data, you focus on good targets with good biology that are predicted to be distributed well and have good safety profiles, those molecules are very likely to get approved and create tremendous value for patients.
And so our view is, I think about the physicians that run trials that are clamoring for those kinds of molecules.
I think about the patients that are looking for more selective therapies.
And so our view is that is the highest leverage application of AI in healthcare, medicine, or broadly.
Just a business question here.
What is the space of things which are the target is the biology is understood, the structure is understood, and that, you know, the only thing you really need to do is find the right lock.
That seems like you probably would have already picked over that class of targets.
In some sense, those are easy targets.
Right.
So how much opportunity is there for that?
So there's two orthogonal concepts.
Known biology is orthogonal to the ease with which one can drug that target.
Sometimes, unfortunately, it seems they anti-correlate.
And that's often it seems that the most appealing targets from a validation perspective.
Seem to be really hard to drug.
Do they anticorrelate or is it just that we have picked all the low-hanging fruit which do solve both things?
I think it is likely the case that many of the, quote, easier targets for a variety of reasons have been sort of, many of them have been drugged, not all.
But even in those cases, people love to sort of have this false dichotomy of easy versus hard targets and, well, this one's drugged versus it's picked over, therefore it's not.
We at Genesis do most of our work through large pharma.
Like most of what we do is providing AI services basically to major pharma companies like Gilead.
And we recently announced our expansion of our collaboration with Insights, which we're excited about.
And I can't give details of specific targets that they're working on, but what I can say is there's a real range from first-in-class chemical matter.
What it means is, We believe this target causes a disease.
There's no known molecules that bind to that target.
We need to find the first binders ever, the true zero to one cases.
But there's a wide variety of targets we work on that are more, I'd say, one to 10 cases where sometimes there is an approved agent or sometimes there's molecules that are only preclinical, but they're not optimal, where if you can improve upon those preclinical molecules and get them to development candidates, they're ready to get into patients.
Or if you can improve upon the existing clinical or approved agents, you can create a lot of value for patients.
I'll give a public example.
We're not working on this target at all, but you just look at the progression of ALK inhibitors, and you look at, to make it very concrete, charts of patient survival.
You might have said, well, when the first ALK inhibitor came out, Why do we need another?
We've drugged elk.
We're done.
But it is a qualitative, not just a quantitative difference, when you compare patient survival curves from first-generation elk inhibitors to later-generation elk inhibitors of how much longer those patients live, how many more get benefit from it.
And so I think there's not only value in first-in-class, there's enormous value for patients in best-in-class too.
And I think...
Both are areas that we focus on, if that makes sense.
Okay.
Yeah.
So we so far have talked about genesis modeling in terms of, let's say, this one specific problem of protein ligand binding.
How does your overall world work in terms of developing this?
I mean, I assume that you must, I mean, I do know that you have other things that you've worked on in solving the broader early phase drug discovery problem.
I keep joking about that when Nobel Prize was given for AlphaFold3, a lot of people thought that like drug discovery is solved.
And it's very far from the truth.
Yeah, you can maybe, maybe, so maybe you can predict crystal structure, right?
At low resolution.
At low resolution.
Well, we're doing better, I think.
Yes.
Assume you can predict crystal structure.
Does it mean the drug discovery is solved?
No.
Obviously not.
The broader community, like protein folding community and drug discovery community immediately said there's other things you care about, like dynamics, like just a single static structure is not enough to understand what's going on, interactions.
So there's lots of things there in addition to just like having a static structure.
I think many people thought those were very useful, like they've really accelerated science.
it was very clearly a starting point and not an end condition.
Right, but if you look outside of people working in the field, like the general popular community, it was a pretty wild, wide belief that the problem is solved.
Now, in reality, besides critical structure...
I just have to clarify that, otherwise, there's an entire ML structure biology community which will all just jump on us and say, no, no, not at all a problem.
Got to throw those caveats in.
Absolutely not a solid problem.
In reality, you need to predict so many other properties, like other properties that Evan mentioned.
Like basically you're designing the key for a log, like that small molecule that sticks to your protein in the body, but you don't want the small molecule to stick to everything in your body because that's probably going to cause issues.
You also want to make sure that small molecule can be soluble so that you can ingest it as a pill.
Those are important properties too.
You want to make sure it doesn't have any other side effects.
And predicting all of those properties is also just as important as, or maybe even more important than predicting the crystal structure itself.
And of course, as a company, and we're very proud of it, we're not just building models for predicting crystal structure.
We're building models for predicting all of those properties and basically enable drug hunters to kind of become hunter techs more effective in their daily job.
I'm curious about this.
The pipeline that you have, you mentioned you have different pharma partners, GSK or whomever you're working with.
Do they have...
targets that were sort of drugs that were designed by you that are in clinical trials, approved?
Where do we stand with all that?
We're limited in what we can say about these are famously secretive, which makes sense.
What I can say, which was recently publicly disclosed as an example, is we just expanded our partnership with Insights and that started approximately a little over a year ago, the initial work together.
And that spanned the two sort of bookends of the drug discovery process in some way.
So one of the programs we worked on together was a case where a very challenging target where chemical matter existed, but there's a key binary event which is getting to a DC or a development candidate.
And that is the molecule or ideally a set of molecules, all of which are possible to be the agent for a phase one clinical trial.
And we had to do some work together where we take our foundational models, our base models, we fine tune them on Insights data and use them in a close collaboration to sort of work together to get to a DC faster.
We're getting substantially closer in that case, which is one of the reasons we're excited to expand our work together.
And on the flip side of that is one of the other areas that worked together was on a protein with also very nice linkage to a very severe disease where there was no known.
chemical matter, like no patents, no papers, in this case, no co-crystal structure of a ligand, another synonym for a small molecule, that bound to that protein.
So we had to find the first ever known chemical matter to bind to it.
And then we'd progress those, what are called hits.
Hits are like the first molecules that bind to your protein, progress those into inhibitors that are active.
called biochemical assays, which are more enzyme-based than cellular assays, which means that in an actual living cell model, your molecule is active as well.
So it was really based on a concrete set of accomplishments together that we wanted to expand the collaboration.
And unfortunately, outside of that, there's really limited any other versions of what we can say.
But the objective of Genesis is to create medicines that patients wish they had.
And the way we'll be able to do that is by working with as many pharma and biotech partners as possible for whom their comparative advantage is discovery, clinical development, commercialization.
And our comparative advantage, of course, is in AI, where we can put those two expertise together.
in a synergistic, not just an additive way and make medicines together that otherwise would not be possible.
So that's the name of the game that we're in this for is those clinical outcomes that you're pointing out.
And I'm very excited to be able to share those as they rise in the coming years.
So one thing that is you guys talk about in your website and whatever is this one angstrom threshold at which.
a protein structure prediction and the binding between that and a small molecule becomes useful.
Can you talk a little bit about like, okay, why is that the case?
How did you do it when others have failed?
And you've spoken a little bit about that with the sort of biases in the model, the synthetic data, maybe there's more.
How does that impact sort of downstream the actual things that we're talking about here with?
you know, the ADMAT and all that stuff.
When we're talking about molecular interaction, the scale of two angstrom that people typically measured is just too big.
So you can think about like in image generation model, you would generate a picture and it's like fuzzy.
So yeah, it's like in general, right?
But you can't discern the details and the details really matter here.
Like with two angstrom accuracy, your entire aromatic ring can be flipped.
And it will still be a valid output.
The worst part is that unlike an image, which is blurry, you don't even know it's blurry.
You flip around an aromatic ring and it looks just fine.
Or even worse, a heterocyclic aromatic ring, and then you're really in trouble.
And it really matters in this case because individual atoms need to establish connections here.
That level of accuracy that they're pushing for one angstrom is really important.
So maybe it's almost like with a lot of generative image models, they will fill in a detail which just doesn't exist.
So if you were to try to do like forensics and you tell, you know, your vision transformer to enhance an image and suddenly it just pops up a face, but that's actually not the person's face, right?
But like you don't know that and it feels just fine.
And so if you're trying to use this as a structural hypothesis as a med chemist, you're kind of screwed.
If you're looking at the wrong face as an investigator, it's not going to take.
Yeah.
Yeah.
I really like your framing.
And then as Evan said, as a downstream, you run a bunch of other models, like physics based models, and all of those little issues may compound.
And then obviously, if you made the wrong prediction, like fundamentally wrong, then the downstream predictions are also going to be wrong.
I just want to say for your question, I really appreciate your mid 2000s TV reference about enhancing forensic images.
So I just want to say I appreciate that before we get back into the details here.
We talked about poses.
So maybe the contrarian take is that a pose isn't even a well-defined concept and that, you know, the best way of thinking about these things is this like, you know, probability distribution over things.
There's a small molecule which probably lives in the binding pocket, something like this.
I mean, is this pose, like this is an abstraction that humans use and not really in some sense ground truth.
So I'm actually almost kind of interested to actually hear that like, oh, we have this one-inch-strom threshold.
This is like really important for med chemists.
I would also talk to some med chemists who would say, I mean, this pose isn't even real.
This is just like maybe a most probable configuration or, you know, how do you think about that?
abstraction in general and do you do you explore conformational space and provide that as tools and yeah how do you interact with just like a single structure versus like an ensemble and sorry i don't know if that question even makes sense but yes it's an abstraction but it's a very useful abstraction it helps us to build up confidence but a particular model output is actually valid uh what you did not just straight up hallucinate at something.
Because yes, ultimately what matters is binding affinity or potency and you can straight up predict that with your model and just skip the entire post generation step.
But then you only have a single number and that number might as well be completely hallucinated and you have no means to validate whether that number even makes any sense.
So as much as pauses are not perfect, they're still a very useful tool for the entire process.
That's right.
Going into that correction, you know, just to give a frame, I'm jumping into a technical rabbit hole here, but that like just because you have a pose also, I mean, there's like entropic and entropic contributions and predicting binding affinity is not just about getting the energy right, but actually is this even likely to make it into the binding pocket and is it like to live there long term?
So that's much more than just you know, affinity, potency prediction is much more than just, is this the right pose, but does this have the broader properties it needs to be a long-lived molecule in this state?
And how do you kind of deal with those things too?
A note for sort of the wider AI audience, probably resonate both with practitioners as well as users, is, you know, obviously something that's blossomed in the past six months is agents.
And we love agents.
Who doesn't?
There's a lot of, I'm going to say a few necessary but not sufficient conditions, I'd say, in my response to a really deep question there.
We all remember what agents were like, let's say, mid-last year.
And let's just say there is positive value and there's negative value, and both can be amplified by agents.
And why is that?
Agents are only as useful as the underlying models that they're orchestrating.
And let's think about coding.
If your coding model even makes subtle but real bugs, your agents are just going to amplify those issues.
And you're going to end up with not only slop, but something that may be anti-useful that might give the user incorrect information.
We all remember what that age was like mid last year that made a lot of people lose patience.
I would say, with claims about LLMs for agentic engineering.
Something changed, like clearly a threshold was met, and even though these models are still not perfect, the utility of LLMs for software engineering are so obvious now.
I mean, obviously it's been a huge tailwind and very useful for us for replacing a lot of the drudgery of coding and getting a focus a little bit more on some of the more strategic issues that matter.
You could draw a direct analogy from that to what we're talking about here in that we obviously know it will be no surprise, but we're working on an agentic platform for 24-7 drug discovery.
You can imagine just fleets of hundreds of med chemists and CAD scientists working nights and weekends all the time for your drug targets.
The codename for that gem is Sapphire.
The prerequisite for that...
was we needed the underlying models for pose, 3D complex prediction, potency, ADME, to all be good enough for an agent using these models 24-7 to create molecules that medicinal chemists would actually want to make and not laugh at.
And if your model is sitting at 1.8, 1.9, Anxstrom, RMSD, that's slop.
Most likely.
And let me be really direct about it.
What you're hearing, so you're hearing bimodal distribution from people of the utility of a 3D pose and how, quote, real it is.
The reality is that for a highly potent ligand, almost certainly there is a large portion of that molecule with a very well-defined 3D position, down to even half angstrom.
And if you don't believe me, you can open up the electron density diagram in the PDB.
It's all available online.
And you can literally see, in some cases, aromatic rings with missing density in the middle.
So it's not just a construct on a blackboard that your organic chemistry professor showed you.
You can literally see a donut, a torus of electron density for an aromatic ring.
However, there's going to be solvent-exposed areas, oftentimes.
that will be less important for binding affinity, but maybe are important for solubility or other properties of your molecule, those sometimes are less defined because sometimes in reality, to your point, they're more dynamic, they're flopping around in solvent.
But the critical piece as an upstream indicator that's valuable for predicting the free energy of binding is to get the core of your molecule that specifically interacts with the protein.
to be correct, to sub-unctrum resolution.
And why does this matter, just to use it intuitively?
So a hydrogen bond, everyone in the audience has probably heard about a hydrogen bond.
It's one of those critical forms of non-covalent interactions.
It's how nucleic acids are held together.
It's how most ligands and proteins interact.
It's how proteins form secondary structures.
Hydrogen bonds have a very specific angle and distance.
And the distances from...
The donor, the acceptor, heavy atom, it's 2.7 angstroms to 3.3 angstroms.
And if I do my math right, that's a 0.6 angstrom gap.
And outside of that, it's not hydrogen bond.
If it's less than that, it's a clash.
If it's more than that, the interaction is much, much weaker pretty quickly.
And angstrom, for those who don't know, is one-tenth of a nanometer.
So drug discovery really is a science of resolution.
If your accuracy is not sufficient, it will therefore not be useful for the downstream things that you care about, which is both potency prediction, but also prospective design.
What molecule do I make next?
So that's why our view is that the history of innovation and startups is that the ones that do best are ones that focus on one.
well-defined but very important problem.
And we think our ability to get higher resolution predictions stems from our judicious focus on small, medium-sized molecule design rather than boiling the ocean.
Yeah.
So that's the what and the why.
So what about the how?
How did you get to 1-Engstrom and sub 1-Engstrom?
I'm going to give you an extremely boring answer.
Okay.
Which I think is actually true for entire AI field.
Like where...
Three things that matter in AI.
It's data, infrastructure, and evals.
Okay.
Right?
So you can only improve what you measure.
And once you are very careful about measuring what matters, and you have really talented people on the team, they're going to figure out how to heal client but measure, right?
So from the start, we actually focused on less than one angstrom precision.
And that led to a bunch of small decisions in the process.
compound right if your team never looks into this metric at all then you will never train the model if it is good at it.
And if you're constantly looking into that, then you're going to achieve that.
So it's the right objective plus good science and engineering.
Correct.
And it propagates everywhere, right through the whole stack.
It propagates through how you look at the data.
Maybe like how do you filter out data?
Some data is more noisy.
So maybe you don't need to see it or maybe you don't need to see it later in the training.
It propagates through your modeling architecture and propagates through your loss.
How long have you been working towards this specific goal?
I'm curious, this is not something that you broadly hear in the community talk about One Angstrom.
People do say to have decided Two Angstrom is sort of the canonical benchmark.
And I've never heard someone say One Angstrom before is the cutoff until like reading things coming out of Genesis.
How long have you decided this is the number we need to hill climb on?
How?
direct as this focus been.
And in like the evolution of the company, I'm just wondering, like, how did this come about?
I think one of our most important secret is that we're working on real drug programs, either with partners or in-house.
And when you actually work on real drug programs, you see the failure modes.
You see what works and what doesn't work.
And it's pretty obvious when you see at the outputs of real programs what kind of failure modes are happening.
And it becomes very obvious that Twangstrom is just not working out for those setups.
Right.
So, I mean, there's a lot of really smart med chemists who are also, you know, who think very carefully about benchmarks and people I respect very much and who also have prosecuted actual drug discovery programs.
So, I know I'm just kind of curious, like, why hasn't this become part of the community?
Is this literally just that the community has never been able to succeed at a winning some benchmarks?
So, they kind of...
settled it to as a, okay, something we can aspire to or I know just.
But I mean, to your point, it sounds like there's a, in my experience with pharma, there are plenty of problems for which people, there's kind of these known things amongst the technical experts in a subdomain and but that it doesn't get out of pharma or it doesn't get the attention it deserves because.
uh you know for whatever some of the information is proprietary and gets kind of passed from company to company but never really released into the public is this your estimation of what's going on have you ever heard of sweet bench so um gemini does pretty well on sweet bench sometimes gemini publishes models that win on some of those software benchmarks raise your hand if you're using gemini to write code right now instead of you know the obvious other name competitors Not me.
No one.
Like, why would you do that?
It's obviously worse in practice.
And I'd say actually that in this case, if you look at the provenance here of how that happened, RMSD less than two came originally from, again, Brent and I have been in the space forever, from docking studies, long before AI models for post-prediction, when physics-based docking studies from academic institutions, because usually large, the proprietary software makers They didn't want to benchmark their methods against other methods.
So academics had to get licenses and try it.
And they introduced RMSD lesson too, which is not surprising.
They're academics.
They're not using these things to make drugs.
They're using it to write papers.
And so that got sort of the provenances that kind of got...
repurposed by the AI community.
But the actual trend is much more in the direction that we're saying.
So the first big innovation there was Posebusters, we'd say, put out by a lab at University of Oxford, which pointed out that RMSD itself is insufficient and we need to look at physical validity as well.
So that's Posebusters that I'm talking about, not the benchmark, but the metric.
So Oxford improved that.
And in the latest release from OpenBind.
We just published our benchmarks at Genesis on OpenBind set.
We'll talk to that in a bit.
But if you look in their, you know, that's their original publication, their default metrics are using RMSD, PoseBusters validity, and LDDT.
So there's clearly an acknowledgement that's already been made that RMSD less than two is insufficient, and the field is now rapidly evolving to acknowledge that.
You think that the academic literature is going to establish some benchmark?
Maybe it's one Engstrom, maybe it's some LDD or some other metric that will kind of converge and then hopefully as a community drive forward these like more accurate modeling strategies.
I would say there was an eval crisis in our field that is now in transition right now as our field, which is...
Previously, let's say a lot more quiet, but if you go to NeurIPS and ICML every year, year after year, their workshops get a lot more crowded.
Those evals are in transition now that we realize the flaws of what came before it.
That was a really interesting point about, you know, pick the right evals and then just do good machine learning like you normally would do.
But I kind of picked up on something Evan said earlier, which is that you started out with, I guess, what you call potential net.
I don't think you defined that earlier, but, you know, this graph-based network.
And then you did a lot of computational simulations kind of after that.
And now you're sort of going back into not once generative modeling took off.
I'm curious about how that evolution worked and how did you find computational techniques to be like really crucial to sort of building on the ML for a while?
And did that sort of computation lead back into generative?
Or did you start collecting data to build a generative, computational data as like, this is a way of data documentation for a generative model?
Like, what's sort of the history of that?
And then what led to those decisions to the extent you can talk about them?
I will say that the last line of PyTorch I've written is much further back in history than the most recent line of PyTorch that Sergei has committed.
So I want to make sure that he gives his opinion as well.
I want to make sure that I'm going to directly answer your question.
I want to make sure that I address something you said like a little bit of time ago that I think kind of got lost, which is you asked about how we do other things, including admin prediction.
So I just want to make a comment about that.
I can spend all day, and I'm happy to dive into details about Perl and the evolution, and we will.
I just kind of want to give a shout out to the fact that it is, one, important pillar, but not the only pillar that matters in the trajectory of a drug discovery campaign.
And you mentioned the PotentialNet paper, which we're happy to see have become influential since you and I were working in the space when it was very much in the future, let's say.
But now it is the future, so it's great.
Another paper we published around the same time was on neural networks for ADME prediction.
And we've published two, actually.
Sorry, ADMAT.
Can you define?
Oh, sorry.
Thank you.
Yeah.
Absorption, distribution, metabolism, elimination, and toxicity are the five prongs that comprise ADMIT.
These are all the properties you talked about before that you just have to get right in order to make a drug successful.
Correct.
So I would say there's over 30 or so assays.
Each of which you could imagine, if you're a neural net person, like a multitask neural network or multi-head, it's got to predict over 30, you know, three dozen-ish properties, each of which, if it's in the wrong range, means your molecule is not a drug, it's just a tool.
And so these are things like solubility, which Sergey mentioned, which is important for formulation, as in can it be made into a pill that you can take orally?
your oral bioavailability, whether or not you're inhibiting certain enzymes called the cytochrome P450s and its different variants.
HERG, a critical channel, which if you inhibit it too much can cause cardiotoxicity.
So these are, it's an alphabet soup of things that most people here haven't heard of.
And a lot of these are extremely hard specifically because it's not like a single causal.
There are oftentimes many processes, many pathways which are involved in defining this one endpoint.
Correct.
And data sets are oftentimes comically small.
Right.
Unless you have, you know, I guess pharma to help you out.
But at least open source, it's quite a hard problem.
It's sparse in the public domain.
I'd say there's actually a range from really directly predictable properties to ones, let's say, that, as you point out, are actually amalgams of other.
So whether or not you're inhibiting CYP3A4, that's really specific.
It's a certain protein that you're inhibiting.
So that's something that Perl, for example, could actually model and expect to see some performance here.
Yeah.
We've been doing 3D work on ADME prediction before anyone else was.
And there was a bunch of work.
that ended up launching the company.
Back in the day when AI could still be in peer-reviewed journals, not just sort of like random white papers on archive, we published MoleculeNet, which has been cited a few thousands of times at some point.
And we also published this paper showing that multitask graph neural networks were the best at a time for doing ADMET prediction on large pharma data sets.
And that paper is one of the most cited papers on AI for ADME ever now.
A lot of papers, as you know, it's like you write them and you kind of move on.
But both those works have been quite influential as well.
So our history from the beginning has been to work on not just one problem, but to focus on drug discovery.
And in so doing, being able to have the bandwidth to focus on building all of the ML models that are needed for drug discovery and not get distracted by the biological discovery processes that are needed.
the target ID side or the clinical trial side, but really focus on all those tasks that are required for drug discovery and also molecular generation, right?
These are all important.
And I just wanted to make the point at the beginning, but as you rightly point out, Brandon, PERL is a 3D structure prediction model and many of these ADMET properties can be posed as those.
So it is therefore useful and not just on target potency prediction as well.
All of them matter, and we've had to tackle all of them.
Do you use Perl for all these tests, or are they focused on some of them?
We'll be sharing some things publicly in the coming period, but I think we've had enough news of results lately, I think so far, if that makes sense.
Every time you publish something, it's a lot of work for the team to put together.
So we'll do that in the coming period.
The most recent thing we published was obviously the OpenBind results, which was a standard 3D prediction task.
So one thing I want to ask about, or I kind of keep hitting it this.
So, you know, you talked about there's a lot of synthetic data and then there's, you know, there's ML modeling, which is very kind of modern, general model flavor.
I want to talk about kind of the feedback between how your computation, your ML, and how wet lab data went into developing this model.
And, you know, we talked about priors.
I think maybe one, it's been typically hard to scale protein ligand models well in a way which meaningfully generalizes.
And there, you know, this recent Pearl paper and some of your other recent results we'll talk about shortly have shown true generalization.
And what I'm kind of curious about is how do these different aspects feed together?
to give you this power.
And specifically, you know, just saying computation is interesting, but like you have to be very careful about computation.
MD has all sorts of biases.
If you're not right, it can give you poor results.
And so I'm curious about how all of this worked together and especially in the history of the development of the company, how did you kind of get to this point?
There's this, the classic concept of the narrative fallacy where in retrospect, it's...
everything seems obvious and you can draw really linear, you know, processes of how we got from point A to point Z.
But the reality is always more interesting and much, much messier and much more nonlinear, I think.
And for Sergey and I, who've been training or taming neural nets for over a decade, we remember a lot of different eras in the space.
And...
We would have been so excited if we could tell ourselves the capabilities of our technologies 10 years ago.
And if we told ourselves how we got there, it might have seemed somewhat obvious, but still there were some things that would have been really difficult to foresee.
So one example is the concept of using generative AI for the molecular space is not a new concept, but the reduction of practice would have been very difficult.
a decade ago.
And so as one concrete example of that, I used to co-run the Stanford AI salon.
We had fun organizing Stanford and the Gates Computer Science Building.
We'd get some interesting speakers and like 25 people having wine and cheese.
And they're all now like some very like just straight up famous people, right?
But back then, AI was a much smaller field.
And I remember...
very clearly in like 2017, 2018, talking about GANs and how generative adversarial networks and how they're clearly the future of image generation, obviously.
And so there was a lot of work then also in, well, can we just apply GANs to produce conformations of proteins or can we try to use them to produce protein ligand poses?
for all the same reasons that those models were really tricky to train for images.
Mode collapse was the most famous sort of problem.
They didn't work very well for proteins or protein-legged systems.
And we sort of had to wait for the right primitive to get created.
And that turned out to be diffusion, which turned out to be a much more useful primitive for the space.
And interesting thing is actually a lot of...
image and video models, some are using diffusion, but some of those have actually gone autoregressive.
And so what's kind of cool is right now for people that are interested in really core fundamental AI research, actually some of the most innovative diffusion research is happening in our field, is happening in 3D structure prediction right now.
No one would have predicted that then, but now that's kind of a pillar of diffusion, I'd say.
That was unpredictable 10 years ago, everything I just said.
And in parallel to that, long before the advent of 3D diffusion models for generative tasks in chemistry, we were building a variety of tools for the problems of drug discovery at hand, some of which were using physics-based methods for predicting potency or even for...
helping predict certain ME properties.
And same with molecular generation.
That is like using different techniques to generate new molecular ideas.
As Sergey pointed out, there's 10 to the 60 drug-like small molecules.
Searching that space efficiently is hard.
So we've been working on that problem.
And those things happened to be available when we wanted to...
take what was then the very nascent area of co-folding, which was clearly exciting, but not at all ready for prime time, not at all ready for what a medicinal chemist would want to use in their day-to-day work, and take it to the realm of useful and in some cases irreplaceable, like clearly superior to a non-co-folding method.
It kind of just happened that we had been building those other primitives.
And so they were available to us so we could put them together to build out, for example, the synthetic data pipeline or to use some of the inference time techniques.
Like we were ready to do those rapidly because we had been working on orthogonal techniques in ways of approaching problems.
So there was some thing here that was contemplative, but like, unless any discovery, like the discovery of penicillin, I'm not saying what we're doing is on the same order, magnitude, or benefit to humanity as penicillin, but there's always this element of if your laser focus is on a problem for long enough and spend enough time in a lab or in our case on a computer banging your head against the wall in these problems, you can create the luck.
as it were, that will enable some of the developments.
How does the lab interact with the development process?
We, as I mentioned before, we train other models, not just structure prediction models.
And for those specific models, lab outputs are extremely useful right away, like potency prediction.
right those outputs can be used directly for for model training um but what i'm most excited about going forward is reinforcement learning i think that's coming in our field and we have seen early signs of uh it working already with our models so where you basically put initially uh maybe physics-based feedback to improve your models through rl typical rl loop But eventually you can go all the way down to the lab envelope setups where your model is not only producing like predictions, but you synthesize based on those predictions and then you measure downstream properties and then you feed back those predictions into that model.
So how do you get enough volume?
I mean, do you have automation to do that or do you like large campaigns that aren't necessarily automated, but generate a lot of data?
One thing I'm super proud of is our partnership with Insight.
They're such an amazing company.
They are extremely, extremely good in producing data.
Basically taking the compounds, generating them, creating them, and then measuring the downstream properties and then sending those results back.
This is such a partnership kind of born in heaven for us between Genesis and Insight.
But we are able to train models, give predictions and have results back from inside super quickly.
So your like sort of rollouts include a lab, like a lab iteration?
Essentially.
Yes.
That's amazing.
Yeah.
And diffusion slow.
So it's about the same speed of that.
It is true.
It's like I'm a big believer that companies are typically.
really good at one or two things and do best when they can focus on it.
And in the same way, we've been just really doggedly focused on developing the best AI models for drug discovery.
Like, Insight has that level of maniacal focus on optimizing drug discovery and development.
And another sort of trend people like to be talking about is sort of, of course, the rise of China in biotech.
It's an elephant in the room.
There's no point avoiding it.
It's so in the zeitgeist right now.
And where so many Western companies have, you know, had become to rely more on CROs to do a lot of their wetland work, you know, Insight became like so state of the art in terms of the experimental capabilities they had in-house.
Their productivity is extremely high.
And as Sergey said, it's a matchboard in heaven for us because what we thrive on is continuous learning of the models.
So we want to have design, make, test, analyze cycles that are as rapid as possible and continuously fine tune, in some cases depending, retrain the models based on what we see in the lab.
And so that partnership is...
One of, if not the first ever, that enables that sort of true joint foundation model training on historical and also prospective data.
And as ML practitioners, we all know here that data is just such a critical input in the whole, critical ingredient.
So it's just extremely exciting for us.
And that's a range, you know, from reflecting our conversation so far here.
from structure, potency and a variety of adnet properties as well.
So I think it immediately will improve the strength of our models and be really powerful for generally accelerating drug discovery.
And just we were joking around about the time it takes to do diffusion, but like, I don't know, I don't know if you can disclose this, but.
How long does it take to, you know, you email them with, here's some, here's, I don't know, how many, 10, 100 compounds, and then they get back to you with measurements of those realized compounds in, what, a day, a week, a month, an hour?
It depends.
Like, some compounds are really easy to knock out, really known, like a very well-characterized reaction where the...
usual conditions just work.
And sometimes it's, oh, this coupling actually didn't work like the literature said it would, and we have to try different conditions.
So it depends.
I'm saying this so the audience understands that for all the claims out there about just like robotic labs, automating synthesis, the reality is just a lot more complicated than that.
Sorry to interrupt.
What are some of the complications there?
We've talked about radical.
So they all do some form of automation.
And again, not trying to get you to get in some fight with them, but what's the contrary take here on, okay, what goes wrong with Automated Lab?
Okay.
So this is also one of the reasons, as Sergey mentioned, we're really excited about reinforcement learning because It circumvents the problems of having to do very fast design, make, test, analyze cycles from the lab to feedback in the model.
Ideally, we can throw more and more GPUs at the problem and have the model self-trained in the same way that we once did for board games in late 2010s or most recently for coding.
That's a key North Star for us.
However...
before anyone tells you that there will be a country of geniuses in a data room just solving drug discovery, there are some problems in our space that don't lend themselves to RL as naturally and that will require experimental intervention.
And so to drive in the specifics of why, to give some examples.
And so, you know, coding was always my North Star, what I wanted to do with my life.
I spent a fair bit of time myself in the wet lab as a very mediocre experimental scientist before I found my true joy back when I got to graduate school and I can just focus on coding all day.
The reality is quite messy.
So in order to make a new molecule, you have to synthesize it.
What that means is different chemical reagents and catalysts.
and temperatures and solvents coming together in the right way in the right protocol to try to make your new molecule.
And once you do that, you need to purify it.
If you don't do that, your positive or negative assay read could be an artifact.
You need to characterize it.
So you need to use NMR and other techniques like mass spec to indicate that what you made in your vial is what you actually sought to make in the first place.
which is not as trivial as it sounds either.
That's not only true for us, by the way.
For anyone doing material science for AI, it's the same sort of challenge where did I make what I actually made is actually not a trivial question at all.
It's easier in small molecule drug discovery than is in materials, but it's still non-trivial and it takes time.
We've seen so many cases where screenings of large molecules sounds great in paper because...
You can test millions, billions of compounds in one go, so you have more shots on goal.
And as a bonus, you have data for your model, right?
You just got millions of data points.
Wow, okay, the reality is that the translation of high-throughput screens, whether it's DEL, DNA-encoded libraries, or more traditional screens, the R-squared of those predictions to, like, the actual business of resynthesizing a molecule de novo and doing a...
low-throughput, high-fidelity experiment is a shockingly low R-squared.
That false positive rate is enormous for a variety of reasons.
And so that's one of the many reasons why people are so excited about AI making those predictions because in theory, they can be even cleaner than a lot of the wet lab work, especially the high-throughput work, and yield to higher enrichment and higher true positive rates.
Why can it be more?
Is it just because the experiments go wrong a lot and even does it matter that you have some sort of very precise robot doing it?
It's still there's too many variables.
There's too much slop in the system.
It's a few things.
So one is that drug discovery frequently involves finding the outliers.
Yeah.
One of the many aspects is so wild about our space is that.
The properties that one is seeking in a molecule very often anti-correlate.
Binding tends to improve the more hydrophobic, the more greasy your compound is because protein binding pockets are greasy.
Guess what?
That makes your solubility worse.
Yeah.
And then you want to improve your solubility often by adding polarity to your compound.
And suddenly, for those that remember high school biology, cells are surrounded by lipid bilayers.
Your molecule might not get through the cell membrane.
because you made your molecule two-polar.
So the properties you want often anti-correlate.
And that's where you get, where multi-parameter optimization of molecules ends up feeling like playing whack-a-mole.
So you're often searching for real outliers.
And that often requires making molecules that are fundamentally novel.
And the kinds of chemistries that today can be automated are fairly constrained.
And so your ability to search chemical space broadly for those really top, top Pareto optimal compounds is actually very limited.
So the benefits you get in speed have a very harsh trade-off with the actual quality and novelty of the molecules that you can make.
So you can do a lot of boring stuff very quickly, is that?
If you're lucky, yeah.
That would be, yeah, if you're lucky.
But to get to the answer to the sort of cutting-edge questions, you really, it's hard to get.
a robotic system to do those kinds of things.
Yes, we would love for that to happen.
It would be a huge tailwind for what we do, but that's not available today.
Not where we're at today.
Right.
And so, I mean, I know, again, not asking you to disclose anything that you can't, but what is Insight's approach that makes them so fast and effective?
I think if one is willing to focus on one problem.
and say no to distractions, it's amazing what can happen.
And I think one is that their talent density is very, very high and they're willing to do what is not universally done now in large or medium sized pharma, which is build so many experimental capabilities in house.
And I actually think that it's possible more companies to try to do that given the global dynamic.
of pharmaceutical development that's changed so dramatically.
Right, so like CRO is taking over everything and now suddenly it's become commoditized, but also your capability to do the cutting-edge experiments goes down.
It can.
It's been a debate for hundreds of years about the benefits of vertical integration.
It comes to all the areas of risk and challenges and upfront costs.
but has all the benefits of controlling the process end to end.
When it works, you can do amazing things when you do things end to end.
Of course, it comes with greater challenges, and I think an appetite needed to do it.
And this kind of brings me to another question I have, again, about more industry.
So maybe now's a good time to talk about your pivot.
So there's name change from Genesis Therapeutics to...
Genesis, molecular AI, and that that I think goes hand in hand with maybe a pivot in the strategy of the company?
Or is that in terms of, okay, we're going to, like what it sounds like to me is we're going to focus on the AI portion and then sell that to pharma rather than be a, you know, sort of a drug company with that also sells the tool.
Is that sort of accurate?
When we started the company, the origin of the company was fundamental deep learning research methods that we developed at Stanford.
And the objective we were trying to solve for was how can we have the greatest impact for patients?
And at the time, the founding of the company was purely in AI research.
There was no real precedent at that time for a pure model company really making it in the realm of just creating value for partners in pharma or really most legacy industries.
And the other, I think, component was the company being founded on AI research.
I think we really wanted to...
show people that we were serious about the biotech domain.
So, you know, I think as part of the origin of let's call us the Genesis Therapeutics, we'll seem a lot more serious to the, you know, all the tried and true folks in pharma, the dyed-in-the-wool med chemists, drug developers who we really wanted to work with.
So I think there is an element of an attempt at nominative determinism there a bit, but also an acknowledgement that...
I don't think anyone had really shown at that time that you could be a pure AI model company and really make it in the space.
And that was, obviously, this was 2019.
So we had the benefit of being real forebears, pioneers in the space.
We had the downside of being pretty early.
Thought to be late, but turns out we're still early innings.
As the company grew and evolved, we really were structured.
kind of like a double helix in that we had extremely strong focused AI research and we were fortunate to be able to recruit some of the most experienced and accomplished drug hunters, like people on our management team and board who have discovered, co-invented, developed many FDA approved drugs.
You talk to most med chemists, most med chemists feel very fortunate if they're able to work on one molecule that gets FDA approval in their whole career.
We're really fortunate to convince these sort of accomplished drug hunters to come join our management team and board at the beginning and steer the show.
So the outcome has always been the same.
Everything we're talking about, it only matters if we end up helping patients and their families at the end of the day.
And we have felt from the beginning that the way to do that at scale is to...
primarily partner with large, medium-sized pharma companies and biotech companies so they can do what they're best at, which is novel biology, clinical development.
We can do what we're best at, which is AI, and bring those two together.
In addition to that, as Sergey has said, there's numerous advantages to having in-house wet lab.
Data generation, clearly.
Dogfooding the models, getting really direct human feedback from real med chemists for what we need to build in an extremely candid way.
And of course, the third thing is that the single most valuable thing in the universe, what I like to joke, is the mirror of Erised tests.
If you ask people, what is the one thing you most want?
And you really thought about it, most people would say a medicine for themselves or a family member.
And so there is...
just immense intrinsic value to having one's own pipeline programs.
And when we interact with pharmaceutical partners, on average, they love that we're working on molecules ourselves because it means that we're not just a bunch of keyboard jockeys, but we live the dream that we are pitching and we need to use these models in practice for ourselves too.
So they're battle-tested in an additional way.
And we're sort of domesticated in that we know what's really difficult about the space.
We're not just sort of selling some pipe dream.
It seems to me like there's been a shift in the last, I don't know, year to two years where suddenly pharma has become interested in buying tools off the shelf.
Whereas previously, there are a lot of companies that started trying to build AI and then...
had trouble.
And so they became pharma companies because that's the only way they could make money, right?
You raise a bunch of money and hope that in 10 years you have a drug.
And that drug...
is successful, then you've made it and otherwise then you fold or like sell off the assets or whatever.
And it seems like now that's changing.
You have a lot of recent AI acquisitions and, you know, in drug discovery and pathology and lots of different related areas.
And so have you seen this shift and is that part of the, are you shifting towards selling models or is this just rebranding just because of that shift?
Like what's going on?
To finish that narrative arc there, the name change was just to reflect what we were really doing in practice, which is we are an AI company as we have been from day negative one when we were coding in PyTorch 0.17 and building the first, when Sergey was scaling transformers on PyTorch, the first one to do that, and we were scaling grapheneal nets in PyTorch, one of the first ones to do that.
So it was reflecting what...
what our identity has been from day negative one, but also alluding to the fact that we are very serious molecule makers ourselves, and we are full stack in that way.
We are also, to be clear, our mission is to create as many medicines as possible as would otherwise not have been created.
And I think the way to get there is to put our technology directly into the hands of as many drug developers as possible, not just our own, but...
in big pharma, mid-cap pharma, and biotech.
And that's partially why we've been building our agents so that our models, which have become so much more powerful, can be easily used and orchestrated in the hands of med chemists, CAD scientists who might have never ever in a line of code in their life.
I find it interesting to hear that you have now started bringing in agents into your development process.
You know, it sounds like from the very get-go, you Your philosophy was much more, these are tools for scientists and you make the best possible tool and this lets them optimize whatever pipeline they're going to use.
But ultimately, this is a tool for MedChem.
And now you are switching and you are saying we are actually going to have automated systems which make decisions and then advance.
First of all, you did allude to you've kind of hit maybe the threshold to, you know, the.
like last November threshold where it just magically became useful.
So maybe that's part of it.
But are there other reasons for going in this direction and like, you know, why?
Why now?
Well, let me ask you a question in return.
How many tools do you think a human met chemist or cod scientist can realistically learn and be proficient in?
Oh, I was going to say, how many can they learn and how many do they use?
Because I've seen that they like to use lots of them, but maybe they're not necessarily good with...
Right, but it's really hard, right?
A lot of those tools are extremely complex.
They have a lot of parameters that you need to set properly and configure them right.
And this is where agents are extremely good at.
They actually know how to use all of those tools and how to orchestrate them well.
Sorry, I wouldn't interrupt you.
When you say agent, then...
Some people like using the term agent to just mean something automated, but other people use agents to specifically mean you have an LLM, which is responsible for orchestrating decisions.
I mean, outside of biotech, it's universally the latter, but sometimes in the space of bio, people are still using agents to refer to any sort of...
decision-based system, which is automated out of humans.
And maybe some of them are even more like traditional rules-based systems or something.
Right.
So in our case, I mean, AI definition of agents, like NLLM, which is able to use all of these tools that we have been painstakingly building over the many years of existence of this company.
And that's a very powerful concept.
Because now as a CAD scientist or a med chemist, you don't need to deeply understand all of the details and all of the hyperparameters that you need to set.
And you can just let this thing go and figure out and solve the problems.
And this is where all of the previous questions that we discussed came up, come back as a very important one.
Agents need to be able to understand what those models are producing.
Why do we care about crystal structure?
Well, the agent can actually look at your predicted crystal structure and make decisions based on that.
Something that we believe is super important to iterate and improve.
You're saying that you can take a crystal structure, which I guess in this case is framed as a graph or like a three-dimensional image that Molestar is interpreting, I mean, in some way, and then it...
it actually can make decisions upon what is binding to what and make hypotheses.
Like you've gotten to that point where you have seen it making effective decisions.
Right.
So there are definitely multiple layers to this question.
You can start with like basic image understanding right here.
Or you can use additional tools to understand what is happening in the crystal structure.
Or you can even train a model which is able to natively understand crystal structure representation and like Image modality is a modality for LLMs these days.
Crystal structure can be a modality for LLMs as well.
I see fine-tune it on sequences that look like that are somehow capturing tokens that are 3D crystal structures.
Right.
Yeah.
So if you historically have treated humans, like you were making tools for humans, so now you have agents.
How do you reimagine the interaction of humans and agents and the decision-making process to go?
Is this you hope to essentially automate humans out?
You want to scale up and do more hypotheses?
Or, yeah, and then also the kind of the skeptic would ask, how do you also tell, make sure that your agents are not making weird decisions where they tweak the wrong hyperparameter and now you give you...
the human a bunch of crap.
You don't even understand the parameters well and know that that happened.
It's a great question and I think the answer here you can see in how coding tools have evolved.
Right, you remember Courser when it just appeared and it was like top top autocomplete.
It was still useful but not to the extent that it is useful right now where I can fire up Courser and it just goes and does everything for me.
So I think we're gonna see a very similar trajectory here.
Now those agents were already useful, but now humans need to provide consistent feedback and basically steer and guide this agent along the way.
But over time, they're going to become more independent.
Still, I don't believe in the full automation, like replacing humans.
I believe humans are going to become way more efficient in the job that they're doing using these tools.
So a human will be providing strategic direction.
to what this agent needs to do.
And the agent is going to be able to go and execute on behalf of a human.
I think this is the best time ever to be a creative human being.
Or really be any human.
Because like the long, long arc of history, starting from when we decided that, eh, hunting and gathering, I don't know about that.
I think I'm going to settle down and do this whole agriculture thing.
I think that started thousands of years of drudgery.
Sometimes you wonder why did we actually do agriculture?
Because everything I've seen seems like agriculture was probably a nut loss on the short term.
It was the only benefit I see in the long run is that if you were a hunter-gatherer and you got sick or born with a genetic illness or got injured, that was like, you're in trouble.
There's not much more we could do for you.
Medicine just didn't really exist.
Ultimately, I think the crowning achievement in my mind to civilization is where the sick people among us in society are not seen as burdens, but people who we want to help.
And that isn't true in the history of the animal kingdom, including humans.
It's a sad reality, but it's true.
Whereas now, because in medicine, we see people who are sick as people who we want to help, who we can help, and we can help them, they could become even more productive members of society, if that's the thing that you care about.
To fast forward that to now, I think that so much of creative work, day-to-day work, if you look at it as a pie chart, time is our most precious resource.
So much of it was being taken up by repetitive tasks that actually didn't require much thought or creativity.
And now it's like we've created this additional headspace for deep thinking.
And for us, you're solving some of the most technically challenging but important problems.
We can use these tools to become wildly more productive and creative.
And that goes to the med chemists, drug hunters, cat scientists who are our direct users.
They can be grand strategists for a drug discovery campaign.
And when you have hundreds of drug discovery scientists working 24-7 on large data centers, just the output you get in terms of discovery is just going to be greater, especially if it's used in the hands of expert scientists who know what they're doing.
We probably should talk about this earlier, but you have some really exciting results from your Perl model based upon essentially a recent open data set challenge that you alluded to earlier.
Sorry we didn't talk about it before, but I would like to give you a chance to talk about this.
So there's this open bind, it was an EVA721A protease, which was a very, very hard target, had a lot of things which made it hard for the community or for co-folding methods to generally work.
And you ran this internally and I think had some, you know, fun results.
Yeah, I mean...
We can also go back to our conversation on evals and what's wrong with evals in general.
Like if you look at the open source models and their performance on public benchmarks, it's kind of very similar across the board.
And what was surprising when OpenBind came up with this new benchmark is that you see a wide variety of performance of the models on that.
And Well, in part of it, it's because it's just a single target.
Okay, I get it.
But it's also because this is the target that those models haven't seen during the training.
So nobody has optimized for that specific target before.
Nobody here climbed on it.
So it was very interesting for us to see how our models are going to perform out of a box on this particular target that we also have never seen before during training or when we developed these models.
So when it came up, we obviously wanted to look into it.
We run Perl system on that.
And we were genuinely surprised and excited about results.
Our numbers are way higher than other published openly available models.
And to me, it reflects how Perl actually behaves in real drug discovery programs.
Because we have seen this kind of...
number differences shifts in internal programs or partnership programs before, but we were unable to talk about them because a lot of this data is like it's either partnership data or internal data.
We cannot disclose it.
Here you have a perfect example where it's an external target, but we didn't know and we just run on it and we see the staggering difference.
So that's kind of what makes me excited.
There are some specifics of this target.
Like it has this flexible loop, but needs to move when you insert your ligand into the right location.
And a lot of methods just are unable to handle it.
And where Perl was exceptionally good is figuring out how to move this loop.
And we are basically correct for every single pose.
Like we are predicting really well this loop movement.
Yeah, so that's the exciting part.
I'm curious about this.
What is it about PERL that allows it to model a dynamic target like that?
PERL model, the way it's trained, right?
It's trained to predict together protein structure and ligand.
And like protein structure and ligand together, they're not necessarily in the same shape or form as protein isolation.
So the whole training process, it kind of nudges Perl into that direction.
So are there lots of examples in the training set like that where there's this dynamic nature where once, during the binding event, the confirmation changes?
There is induced fit throughout the training set.
Yeah.
This is probably inherited from the physics-based simulations that you're using.
It certainly helps.
I think the other component that Sergey alluded to was...
We haven't just been benchmark maxing on the same.
We obviously publish in Runs and Poses.
We published the state of the art numbers in Runs and Poses last year when we published the Pearl technical report with NVIDIA at GTC at the end of last year.
We used Runs and Poses.
We used Posebusters.
Pearl does the best on those benchmarks.
We did that.
However, as Sergey is alluding to, the gap from Pearl's next best model is even larger on the partner drug targets that we work on.
And what that's in part a reflection of is the objective of our company is not maxing benchmarks.
The objective of the company is creating value for biotechs and pharma companies that are usually working on hard drug targets that are different from what's in the PDB.
And so we've had to build the models in a way that are able to extrapolate.
And the open buy-in challenge was just the latest example of that.
Before we wrap up, we always ask two questions.
One is, if you could remove a bottleneck from your industry by fiat, right?
So this is not like for inputs, outputs, business, what would that be?
I can say where my main bottleneck right now is, it's in GPUs.
Okay, everyone says.
Yeah.
Look, GPU prices are going up and up and up.
LLM companies are kind of sucking up all of the GPU capacity out there.
And I firmly believe that what we're doing is genuinely important for humanity, like discovering new medicines for yourself, for your loved ones.
It's something that we all need at some point in our life.
So I think it's very important to unblock that GPU bottleneck for drug discovery.
Is Anthropic actually throttling science because of GPU purchases?
I don't know if you can ask this, but are you looking at non-NVIDIA?
I mean, I guess you guys are NVIDIA partners or whatever.
So, but like they're, I've seen that some companies are pivoting towards a multi-architecture model so that you can take advantage of other suppliers because of the shortage.
NVIDIA has been...
a great supporter of us.
They've invested twice in Genesis and we collaborate closely, like, you know, not just in terms of the capital, but we optimize kernels together.
You know, we cope.
Pearl was co-authored by Genesis and NVIDIA.
They publicized the OpenBind results with us.
We've been very grateful for that.
They're a very, very, very big company.
And, you know, we're not the scale of...
customer, like a meta or something.
So we're grateful that they've given us time and attention nonetheless.
But I think part of that is a self-serving aspect here, which is that obviously no one can time the market, but the amount of hype investment in the traditional LLM space for chatbots, coding sort of thing, at some point...
the amount of alpha will not be what's desired compared to the amount of investment.
And I think regardless of the vicissitudes of the economy, medicines will always be high in demand.
And I think whether it's a desire to help humanity or a self-serving interest in looking for the next big thing, I do think that chip makers...
including NVIDIA, are going to want to get a lot more invested in life sciences because it will always be high in demand and the amount of alpha left in pure LLM space is just getting a little questionable.
I mean, all the LLM companies are starting looking into life sciences right now anyway, right?
GTP Roslyn, there's the Codd Initiative, there's lots of startups and whatever.
Yep.
We think those are all great tailwinds for us.
And then what about you?
What's the button on that besides GPUs?
Oh, I had the same answer as Sergey, for sure.
If I could wave a magic wand, I'd love to have just an enormous H100, H200 cluster.
That would be amazing.
And then the second question we ask is just, do you have a call to action for our audience?
So meaning AI engineers plus scientists get involved with something, come apply for jobs.
do like, what's their call to action for them?
We're definitely hiring.
We're definitely looking for top talent in the space.
And I personally, I transitioned from LLM space to drug discovery.
I've never done drug discovery before during Genesis.
I find this field fascinating and extremely interesting.
You know, in LLM companies, you often find research scientists excited about working on architectures.
but like kind of forced to work on boring stuff.
And honestly, LLM architectures are relatively boring.
I don't know, I probably alienate half of your audience.
I think they probably think the same thing.
But it's like, it's a transformer layer in the end.
Like paper was published in 2017 and you go to any LLM lab today, you will see very, very similar pieces in their models.
Yes, there is a little bit of flavor of MOEs, but otherwise, architectures are fundamentally very similar to what was discovered back in 2017, right?
Well, our architectures and our models are actually very different and very, very interesting to work with.
So if somebody is excited about working on architectures, well, this space is actually...
very, very interesting place to work.
Yeah, I'd like to comment.
Not only are they very, your architecture is very different from the LLM space, but you're actually, you're also very different from what a lot of the other bio ML space is doing as well.
Like it's, it's sort of unique both in their subdomain and more broadly.
We saw as a recent guest, you had one of the, the, the progenitors of ESM fold.
Yeah.
Yeah.
They cite our, model architecture paper that we published in open sourced a few months ago.
So the work that you do here at Genesis, as an AI researcher, there's lots of opportunities for publication, open source, and influencing the field beyond the obvious ways to influence the field by working with top customers that are actually making medicines that change people's lives.
So it's both incredibly intellectually engaging and a lot of opportunities to influence the field and get your name out there.
But if you want, you could also just stay being a cog in the machine of an LLM company.
No shade, no shade.
All right.
Well, thank you so much for coming and speaking with us.
I've personally really enjoyed this.
Hopefully the audience as well.
Evan, Sergey.
Thank you, and we look forward to seeing how things develop, and hopefully you get some good job applications out of there.
We love the show, so we really appreciate taking time to chat.
Yeah, thank you for having us.