Episode: 85 - Daniel Chen on Building Multi-Target Drugs with Logic Gating

May 12, 2026 | When it comes to drug discovery, either you find a rare, clean target and everything clicks, or you spend a decade chasing biology that refuses to cooperate. Daniel Chen, M.D., Ph.D., founder and CEO of Synthetic Design Lab, makes the case for a different playbook: objective-driven protein engineering, where we start with the outcome and design biologics that can execute it with built-in logic. With host Andrew Bradbury, Chen unpacks the SYNTHBODY platform and what “multi-tier logic gating” means in real drug design. They dive into logic-gated antibody drug conjugates, where target summation can boost binding, internalization, trafficking, and payload delivery and where an additional AND gate can trigger multiplier functions that go beyond cytotoxic payloads. 

Links from this episode:  
Specifica
Synthetic Design Lab 

Transcript

Welcome To The Chain

Announcement 0:01

Welcome to The Chain, the podcast exploring the lives, careers, research, and discoveries of protein engineers, scientists, and biotech professionals. We look at the impact their work is having on the field and where the industry is headed. Tune in to stay up to date on the newest advancements and to hear the stories that are impacting the world of biologics.

Andrew Bradbury 0:25

Okay, hi. I'd like to welcome you all to this issue of the chain. My name is Andrew Bradbury. I'm CSO and co-founder of Specifica, which is now an IQVIA business. And today our guest is Dan Chen, CEO and founder of Synthetic Design Lab. And why don't we start? Dan, if you can tell me a bit about yourself and how you came to Synthetic Design Lab.

Daniel Chen 0:47

Thanks so much, Andrew. It's really fun to be here with you and have this conversation. So my journey to Synthetic Design Lab has actually been a very long one. So trained as a medical oncologist and immunologist as a physician scientist. I've worked in many different spaces and related to drug development and engineered therapeutics. But most people probably know me from my work in cancer immunotherapy. I was part of the many people that worked to develop checkpoint inhibitors. I led the development of T centric while I was at Genentech and Roche. And I also wrote the cancer immunity cycle, which I think probably a lot of people in the field know well. But my fascination with engineering and how it plays into the future of therapeutics is probably, I don't know, 10 or 15 years at this point in time. And and was really driven from the thought that, you know, how long can we sustain a field that is essentially discovery focused, right? So when you think when we think about what what are the great drugs that have impacted human health over the last couple of decades, there they tend to be drugs that are based on the discovery of some really fundamentally important target in the human body. So whether we think about VEGF or PD1, PDL1, GLP1, and the tumor targets like HER2, they're discovery focused.

Daniel Chen 2:31

And when we try to find the next one in that line, it often is really difficult, right? Well, what is the successor for PD1 PDL1 or HER2 or VEDGEF? And it it really demonstrates how while discovery-based biology is really important and it was it it really anchored the last era of innovation in in medicine, that it's it's unpredictable and potentially unsustainable. We just don't know how many super important targets there are in the body, like those targets left. I think we'll never be at the end of discovery, but I do think it's a funnel, right? Like we've captured the low-hanging fruit and and even the probably the layer of fruit above the low-hanging fruit. So it's I I would never say that it's it's done, but you know, the human genome project was what, 2001? That was a long time ago. And think about the armies of people that had scientists that have sifted through all of that data for the last 25 years, right? So it it it almost, you know, need is such an important part of the way different fields move. And I feel like the need is so desperately there to move on from just discovery-based biology.

Why Discovery Alone Doesn’t Scale

Daniel Chen 4:02

And, you know, I think what I mean by that is when we look at all of the different biology, all the different targets in the human body, the ones that become drugs are 0.000001% of the of the targets that are out there. I may have missed on my number of zeros there, but it's a very tiny, tiny percentage of of all of the targets. But we know that there are lots and lots of proteins in the human body that are important. They're just not important enough to make a great drug when you just block it.

Andrew Bradbury 4:40

What you're saying is that discovery up until now has involved targets which are almost sort of like yes, no, plus minus. And what's what we're now looking at, and I think if I'm understanding correctly what the Synthetic Design Lab is looking at, is this idea of looking at the complexity of the targets that we have now and seeing how we can exploit that.

Logic Gating For Protein Drugs

Daniel Chen 5:01

Absolutely. Start with, you know, a an objective, right? We are trying to kill a certain type of cancer cell using a certain type of mechanism. Or it could be, you know, we want to help people lose weight, but not lose muscle mass. What whatever it is that we're trying to achieve, set that as an objective and have a way to design drugs to do that, right? That's sort of what happens in other fields, particularly high-tech or engineering. You have an objective. You're trying to build, whether you're trying to build a bridge or a computer or a laser beam, you have an objective, and then you have the tools to try to build that. But in our field, be probably because biology is so complicated, it was so much to learn, that discovery period of biology really stretched out a very, very long time. And so, you know, we're what, a hundred years behind the high-tech field? But I think we can close that gap.

Andrew Bradbury 6:05

Okay, so tell me, explain for our listeners then the SYNTHBODY platform, how it differs from what else is out there.

Daniel Chen 6:14

Yeah, so taking a step back then for us, with an understanding that we wanted to try to create a platform that would allow for more powerful approaches to therapeutics, we formed Synthetic Design Lab and our platform, SYNTHBODY, to try to do just that. And we took a page, again, from the field of high-tech, right? We and we know from 100 years of high-tech engineering that there were certain things that allowed us to create more powerful tools, machines, in our case, hopefully drugs. And one of the most powerful approaches to that is this concept of logic gating. Now, in our field of medicine and biotechnology, we have started to play with very, very simple logic gates, right? So in the field of biologics, you're starting to see these and gated therapeutics that allow for better safety because you tie a clean target to a target that is less clean, and you essentially create the need for both. But that's really just dipping your toe in the water. And we felt like at Synthetic Design Lab that to really push the boundaries of what we can do in medicine, that we would want to create something much more. And so we created this SYNTHBODY platform with the idea of being able to create advanced multi-tiered logic gating that would allow us to not only create better, safer drugs, it would also allow us to create much more efficacious drugs.

Daniel Chen 7:59

And it would even allow us to potentially ungate different ways to achieve your objective based upon these logic gates. And so to do that, we created a platform that has between, these are protein-based biologics that have between six or 12 different active domains, right? These active domains can be binders, but they don't have to be just binders. They can, they're there to achieve an active biologic outcome. And we've learned to create this platform in a way that essentially that the way the the different binders are structured create themselves a logic gate. And that those logic gates are obviously not based on silicon or or actual mechanical switches. They're based on biophysics and they're based on the behavior of proteins at the surface of the cell.

Andrew Bradbury 8:58

So this sounds all very, how can I put it sort of vague and theoretical. I mean, give give me a concrete example of a logic gate in the protein therapeutic world as you see it, so that our listeners can understand.

Logic-Gated ADCs And Target Summation

Daniel Chen 9:13

Yeah. So when we started the company, it was a very much a vague dream because it was it was a hypothesis, right? We had a very clear hypothesis on what we were trying to do and how we were going to achieve it. At this point, I would call it a reality. We've been able to make these molecules, they work really well, they are much more powerful than traditional protein-based biologics, at least in the testing that we've done to date. They have drug-like properties, they're not difficult to manufacture once you design them well, they behave well in animals. So for us, the next real step is obviously the clinical testing. But to your point, like what does this actually look like? Well, what we did is in creating a fundamentally new technology platform, we wanted to limit the amount of initial biologic risk. We wanted it to work. If the technology worked, we wanted it to work well as a drug. And it was one of the reasons why myself, as someone who, you know, has spent so much of my career working in cancer immunotherapy, didn't start with a cancer immunotherapy approach. I think the SYNTHBODY platform will be incredibly powerful for cancer immunotherapy applications, but there's a lot of biologic risk right now in novel cancer immunotherapy because there's just a lot of complexity to what is really leading to immune escape in patients getting cancer immunotherapy. So, what we wanted to do is say, look, let's focus it on something very clear with limited biologic risk. And so to do that, we focused the platform initially on antibody drug conjugates, right? With the idea that we understand they work, we understand that cancer cells can express a whole bunch of different types of proteins on their surface, and that if we can design a drug really well using these advanced logic gates, that we would be able to deliver a lot of our payload into the cancer cell.

Daniel Chen 11:25

In the case of this most simple antibody drug conjugates, those would be cytotoxic payloads. And that would lead to greater, much greater killing of that cancer cell, and you would have a, you know, a great antibody drug conjugate. And so that's how we started, and that's where we're currently primarily focused is being able to build these as logic-gated antibody drug conjugates. And and the way we do that is we look at the the target space on a given cancer cell in a given indication. We essentially like to bucket them into different classes. So we call type A and type B targets. So type A are your great classic antibody drug conjugate targets, right? So obviously HER2 is is the absolute best target, and that's why drugs like in HER2 are st so incredible as clinical drugs. There are there are lots of other targets in the in the type A target space that are clean, but they tend not to be expressed at at high enough levels to drive the type of activity we see with it with HER2 targeted therapy.

Andrew Bradbury 12:44

So the whole the the real issue about HER2, why it's such a great target is just because of this issue of expression, right? I mean, it's just so highly expressed on cell services and nothing else comes close. Is that is that a reasonable assessment?

Daniel Chen 12:57

I mean, I think it's reasonable, but it's also clean, right? Because there are lots of other targets that are expressed at similar levels to HER2 and HER2 High, but essentially zero percent of them are clean the way that HER2 is, right? So that's a huge problem. The and of course HER2 internalizes well. There are there are some targets out there that just don't internalize very well. But I think the major thing that makes HER2 so special is very cancer specific. So you just don't get very much off on target off-tumor tox, and and it's expressed in HER2 high patients at somewhere on the order of 10 to 100 times higher levels than the next best type A target. And so that's a problem. So so one of the ways we think about building these logic gate SYNTHBODY constructs is that we take a look at the type A target space, and those become our summation targets. So these very clean targets can be summated to essentially create a model an antibody drug conjugate that behaves like you know, it thinks it's seeing much more target than it actually is, because the target is no longer one target, it might be three different targets.

Daniel Chen 14:25

So we call that an and better gate. It means that we can target any of those targets. If one of those targets is expressed highly, the antibody drug conjugate will work. But the real magic starts to happen when two of the targets are expressed or three of the targets are expressed, then you have a really superior profile where we see not only very strong binding strength to the targets, we see very strong internalization, we see very fast internalization, we see excellent trafficking within the cell and payload delivery. And then we end up seeing great in vitro and in vivo efficacy. So that's I think that's the starting point is this concept of and better gating to multiple good ADC targets. But my favorite part of it is actually the back end of the logic gate. So in the back end of the logic gate, we can start to build in a AND gate, right? So now you you need to bind to these very specific targets on the surface of the cancer cell to trigger the AND gate. And usually what we have behind the AND gate is what we call a multiplier, something that really then boosts the activity much higher. So beyond just summation.

Daniel Chen 15:49

And, you know, I would say the most intriguing and the most fun part of that is we can build in things that are not necessarily directly just limited to antibody drug conjugated payload delivery, right? We can, we can make that multiplier something that greatly enhances targeting internalization, payload delivery, but we can also build it in to do things like modulate the cancer immune response, or drive chemosensitization, or have an anti-angiogenic component, or and the list goes on and on. There are a lot of things that we can put behind there that then really greatly increase the activity of the SYNTHBODY.

Andrew Bradbury 16:32

So the SYNTHBODY has essentially two AN gate components, right? The first and gate component is a targeting one, and then you're saying there's a second and gate component that is a multiplier that will take a take a targeted SYNTHBODY, or if the SYNTHBODY recognizes a particular cancer cell, and then will actually increase the biological efficacy of the molecule that you're providing. Is that is that reason is that correct?

Daniel Chen 17:00

Yeah, I mean, I think absolutely that is correct. I think one of the things as as we enter new spaces, you know, terminology is really important. And and so I wouldn't call it too and gated. I like to use the term and better. Some people like to use or as that first gate. You know, you could hit either her two or you could hit trope two as an example. But in reality, to me, and better is a is a better logic label. Because what it essentially says is yes, if her two is the only target that you see there, you still work really well. If trope two is the only target you see there, it works really well. But when both of them are there, you get even more activity. And so that's why I think and better is a nice term. But if you want to, you can also say in a multi-tiered logic gate, the first gate is an OR, and then the second gate is an AND, and then the third gate is an AND safety gate.

Andrew Bradbury 18:02

Right. Okay. Well, turning to the AN safety gate, so it's it's been shown that about 30-35% of monoclonal antibodies have off-target binding. And so in a situation where you have multiple binders on a molecule, and let's say you have three, then if you take let's say it's 30%, then 70% times 70% times 70%, brings you down to about only 35% of your molecules are actually going to be specific for the targets that you think they are. Is that something you've looked at?

Daniel Chen 18:33

Yes. So obviously that would be really important, right? So one of the core things in a molecule like this that you get to design and build around is the relationship between affinity and avidity. And obviously, folks familiar with the PEGS conference and are working in the in this field, this is not itself a new concept. But I would say when we're when we look at the lens of biologics through an IgG molecule, your ability to modulate affinity and avidity are very limited, right? You just don't have very much real estate. And so one of the major emergent properties of scale that we get from the SYNTHBODY construct is when you go from two binders to six or 12 binders, you just have so many degrees of freedom to further attune your molecule. And so one of the things we think a lot about is in traditional biologic-based drug development, you know, we generally have favored pushing into very high affinity molecules to get the type of behavior we want out of the biologic.

Daniel Chen 19:47

When you have a SYNTHBODY, you have so much avidity that you actually don't want that, generally don't want astronomical affinities as part of that. Because I think that's where you start to get into this case. If you have really high affinities and really high avidity, you have to be so careful about what you're hitting because of that. You know, it's like that's like a multiplier effect as well. But when you have the ability to create these molecules with a more modest affinity and play to avidity, you actually have a much finer dial to be able to control on-target off-tumor effects. And we and this is something that I have worked on even previous to coming to creating synthetic design lab, this ability to use avidity to tune how your molecule behaves. And so this is something that we just build in very early in the design of our molecules. And throughout the entire discovery to optimization process, we obviously are making sure that we don't see on target off-tumor effects from the affinity and avidity of the binders.

Andrew Bradbury 21:05

Right. So what what are in a SYNTHBODY? So what are what are your binders? Are they are they fabs? Are they VAKs? Are they receptor ligands? What are you using for your SYNTHBODY binders?

Daniel Chen 21:17

Yeah, so we started by building the platform off of protein-based as opposed to antibody-based biologics. So we started in a space where we created molecules with peptide ligand sequences, right? So essentially the ligands of certain receptors we were interested in. But certainly one of the major classes underneath protein biologics is this ability to utilize antibody binding domains as well. And we've generated a lot of our data with antibody binding domains. We can use any type. We've tried all sorts of different types of antibody binders. So certainly we can use Fabs, you can use single-chain FV. I would say that at Synthetic Design Lab, we have a preference for single domain antibody domains, so VHH-like domains. Just because when you're building a complex molecule, you know, there's a lot of complexity already built into it. The ability to limit some of that complexity in a very dynamic, flexible format, I think that's where VHH-like domains can really excel. But we don't think we're limited to that. I think there's a future where, you know, mini-binders and all sorts of other approaches here can be utilized. But we like to focus a lot of our work around single domain antibody regions.

Andrew Bradbury 22:46

So just so our listeners understand, a SYNTHBODY is a bit like an IDG with a series of VHHs tagged at either end. Is that right?

Scale Effects And 3D Cell Geometry

Daniel Chen 22:58

Yeah, I think that's a relatively clean description for where we are today. I would say that the platform is a little broader than that. The platform essentially incorporates six to twelve different binders and in its current iteration in a per specific geometry that we think is favorable. And our molecules include a core region that provides a certain amount of rigidity and other things that you might want as part of a molecule. So in the types of SYNTHBODY constructs that we're currently focused on, that core position is taken up by an engineered FC. And we like that in 2026 because we just haven't seen a different way to get the type of pharmacology that you get out of molecules that contain a FC-like domain. Particularly things like half-life, obviously, are very difficult to modify to a point where you have something that has the pharmacology of an IgG without actually having an IgG. And I think that probably speaks to the complexity of that whole CH2, CH3, FCRN interaction.

Announcement 24:16

Are you enjoying the conversation? We'd love to hear from you. Please subscribe to the podcast and give us a rating. It helps other people find and join the conversation. If you've got speaker or topic ideas, we'd love to hear those too. You can send them in a podcast review.

Andrew Bradbury 24:33

And so when you say six to twelve binding modalities, I mean, is that six different and twelve different binders?

Daniel Chen 24:42

We like to use a minimum valency of two. We don't have to. There will be certain indications where you actually want a minimum valency of one. There'll be some indications where you want a minimum valency of six. But for most of the work that we've done, we like a minimum valency of two. And that minimum valency of two is part of what creates your it creates that and better gate. If we had a minimum valency of one, you would create an AND gate. But by having a minimum valency of two, the molecule can still behave very similarly to a traditional IgG if there was only one target expressed on a cell.

Andrew Bradbury 25:31

There's a good reason biology chose a dimeo for immunology, right?

Daniel Chen 25:35

Well, you get that log order step up with each binding, you know, your valency from one to two, two to three, three to four. But obviously the biggest, the biggest impact of log order increase is when you go from one to two.

Andrew Bradbury 25:49

One to two. Right. Absolutely. So tell me, how how is this different from traditional biomulti-specifics? Yeah, I mean somebody listening to you might might think, well, this just sounds like what we've already heard about. I mean, why why is this a different platform?

Daniel Chen 26:03

So there are a number of reasons. So the first one is just what we think of as scale. And we actually refer to it as emergent properties of scale because they're things that people wouldn't necessarily assume right off the bat, right? So if you think about, okay, well, you know, traditional antibodies have two binders. Obviously, by specifics, go to two different binders with a minimum valency of two. And then we're seeing, you know, a lot of molecules that actually have now four binders to two different targets. That's an option, right? So there are different ways to play this. Genentech and Roche, we, while I was there, we pushed ahead some cool platforms that have this two plus one format. So you had the avidity for the target, but only a single binder for the C D3. So there's all sorts of things you can do there. But when you go to six to 12, you get these emergent properties of scale. Now, the simplest emergent property of scale is just trade-off decisions. So when you look at traditional approaches, if we say, hey, we could add something that does X, right? Well, when you do it in a traditional format, adding X means you're not, you don't have space for Y or Z or something else. On a SYNTHBODY with so many binding domains, you don't generally make those trade-off decisions.

Daniel Chen 27:33

And we like to make our molecules with a number of binders that match the number of things that we think are important to achieve on for a given drug. So that's the easiest one. The ones that are more complicated about emergent properties of scale and how the SYNTHBODY platform differs from other platforms relates to architecture, geometry, 3D spatial effects, and biophysics at the surface of a cell. So in traditional approaches, you just don't have the ability to control very much in that interaction. So, you know, you think about how you anchor anything, right? If you only have one anchor point, you you don't have a lot of degrees to control how you anchor to that point. When you have three, four, five, six different ways to anchor, you suddenly have a lot of design space on how you create the biophysics that happens at the surface of that cell. And that's a big part of the magic. So to us, scale matters. It is not as simple as saying, oh, can't you do X and Y or X, Y, and Z? It has to do with being able to create architecture that can match a certain 3D spatial geometry that we want to achieve on the surface of that cell. And we just haven't seen the ability to do that with more simple structures.

Andrew Bradbury 29:11

So is it fair to say that other sort of traditional bind multi-specifics are essentially constrained by the original molecule, whereas you're looking more at the target and how to best attack it, and then building molecules that have many, many more potential functionalities on them?

Daniel Chen 29:30

Yeah, let's put the multi-functionality at the back end, right? Because that's a nice, we have a cool way to add that on. But the 3D spatial piece is really important. We think of it as the ability to organize communities of proteins on the cell surface. And the types of things we think about, obviously, the simple ones are specificity for your kit your cell, how much is going to be there? What is the proximity between, you know, target A, target B, and target C? Are they close together? Are they far away? Are they excluded? Are they interacting? What is their epitope space in relationship to each other, right? Because if you're going to start to talk about 3D space and mapping 3D space, you have to think about where your epitopes are. And so there's just a lot that goes into how to create powerful 3D mapped logic gates. And and you need a structure that allows you the ability to design around that complexity.

Andrew Bradbury 30:40

So is that so that brings me to my next question, which is is is the way you end up with a SYNTHBODY, is it a screening process or is it an upfront design process? So do you say, okay, I want an antibody that binds target A at epitope one, target B at epitope two, or do you just take the output of an immunization or a selection and just screen all antibodies recognizing target A, combine them with all antibodies recognizing target B and just seeing which one works the best?

Daniel Chen 31:09

Yeah, so we really use a combination of both. We don't think this works well without being able to think, being able to utilize both of the types of things you just mentioned. So on the front end, we think hard about human biology, right? We think about what is it that we're trying to achieve and how do we best achieve it? What are the right targets for that indication? What is the, what are the what do we know biologically about each of the potential targets that we would consider? And so that is the hypothesis-driven piece. So we look at them, we think about where and how we would want to create this if we were perfect in our ability to to do that using what we know. But as you can imagine, the amount of variables that go into something like this is enormous. It is log order at every every binder you add is a log order increase in the complexity of what you're creating. And so the idea that we as scientists understand this well enough that we can that we can just perfectly design something like a blueprint and draw it up and make it the next day is is would be overestimating our own capabilities to understand enormously complex biology.

Daniel Chen 32:29

So what we do is we limit the space by by thinking about it from a scientific standpoint, right? Human biology, engineering, biology all together. But then we we've created the SYNTHBODY engine that allows us to integrate discovery and molecule optimization. And through that process, we essentially can look at massive numbers of variants to try to sort through what it is that that optimizes the what we're trying to achieve. And then we're able to use those big data sets to then pick best molecules to move forward. So it really is using both. And, you know, that screening part not only recognizes what we can't comprehend in this enormous level of complexity for all these layered variables. We do think that machines are going to be really great at sorting through these variables. So part of what we're doing is, you know, we've created a technology to achieve something. We're first to understand a lot of these principles, and we'll we'll use these to create what I think will be fantastic drugs, but we're also using these massive data sets to then help feed the algorithms of the future. And I think in the future, whether we think that's three, five, or ten years away, I think machines will be able to better understand the amount of complexity here when you have this many variables. After all, that's what they're particularly good at.

Andrew Bradbury 34:07

So and so regarding that, I mean, I would see each so each target, I think, has two levels of complexity. One is the affinity of the binding interaction, and then where where actually you're binding. And so if you have two targets, then you you've almost got a forelog issue where you know you can you can bind at 10 different sites and 10 different affinities at each site. I mean, it all becomes incredibly complex. So do you do you decide up front what affinities you want? Or or or again, you have a range of affinities that you want. You know that target A, target B, you don't want to bind too tightly because they might occur together on on normal on individually on normal cells. And so, you know, how how how how do you address the affinity issue?

Daniel Chen 34:46

Yeah, so you know, you start talking about those layered variables and it just starts to make your brain hurt. Absolutely. I mean, it is the the amount of complexity is just enormous, right? And I think part of the type of work we do is we have to have a certain amount of courage and not be overwhelmed by the amount of complexity that you introduce here. I think that as scientists, you know, we tend to be reductionist, and so it's very hard when you suddenly go to the other end of the world and or the spectrum and and think, okay, we're gonna create something that has so many variables that you can't possibly understand it. So the way we tackle it is again through this idea of funneling. So we think about the targets, and then we, as you mentioned, I think the approach of having a range is a good one, right? So you sort of have a sense of what how you want these things to fit together at the beginning. But but there's no point in being precise about it because well, like what precision do you need when you have so many variables you can't understand them all, right? So you just you just know that certain things need to be in a certain range and others can't be, right? So the some of these things that we build in, you know, from the very beginning, there's no way in the world that you can use it in such a molecule without having it behind an hand gate. And so we know that from the beginning and we design with that in mind.

Daniel Chen 36:16

But I will say that's that there are so many interesting and surprising findings that we get out of creating these things. And insights, honestly, it's one of the most fun parts of this work is the insights that you gain that are just unexpected. Like I didn't think it could work this way. And so even though we spend a lot of time up front describing and thinking about and designing around how we want it to behave, we also spend a the inverse of that at the back end, which is wow, I don't understand why it behaved like that, right? And then sometimes it's good and sometimes it's bad. And when it's bad, you're like, oh, that's unfortunate behave like that. Let's figure out why. And when it when it goes in your favor, it's wow, that's super exciting. It worked like that. Let's try to understand it so we can recreate it the next time in a different situation.

Andrew Bradbury 37:14

Can you give a couple of examples? One one good good good insight on a bad one.

Daniel Chen 37:19

U nfortunately, I there's only so much I can say here. What I will say is that geometry in 3D space is you know, has a lot of variables when you link them together like this. And there are just going to be certain things that relate to how molecules are oriented, what their distances are, how they differ between a cancer cell and a normal cell that you can play to. And if you have the ability to create complex molecules and lots of them, and in high throughput, look at how they behave on normal cells, on cancer cells, you know, on cancer cells with different levels of expression, you get molecules that essentially have very different profiles, right? Some molecules are going to hit cancer cells and not normal cells, some are going to hit both, some are going to be very dependent on a particular relationship between target A and target B that you just couldn't have predicted up front, at least not today.

Andrew Bradbury 38:30

Right. I yeah, I know, right? I know Andreas Proctham had some very elegant work showing that it wasn't enough to bind target A and target B, but where you bind target A and target B has enormous effects on the downstream biology of the of the molecule you've designed. So I I guess that's something similar to what you're you're saying without saying.

Daniel Chen 38:51

Yeah, I mean it's so I actually have a publication in immunity from January of 2024. That is the title is Immunity as Biophysics on the Surface of the T cell. And in that piece, one of the things I described is what I called receptor biophysics variables, which I think is somewhat what you're describing here, which is we think about targets on the surface of a cancer cell as being very simplistic, right? It's a nice cartoon to say, hey, you know, you have the target, it sits on the surface of the cell, and there's either a lot of them or not very much of them. But obviously, there's so much more complexity to that. How close together are they? What orientation are they in? Is it important what orientation they're in? How much, where's your target relate in relationship to this cell surface? Is it distal? Is it proximal? Remember, our molecules, as they interact with cells, it's like it's not like a clean cartoon. It's like going through a kelp forest, right? Like there are so many interactions in there.

Daniel Chen 39:56

What is the orientation of your molecule ultimately? Is it face up, face down, sideways? There, there are just so many interactions. And one of my favorite ones is even, okay, we draw these things like stiff sticks on protruding from the surface of the cell. But do they really exist like that? How much bend is there in these molecules? How much bend is there dependent on the epitope you pick, right? Because if you pick a very membrane proximal epitope, maybe there's not very much bend because you're there's you're hitting just a nub on the surface. But if you're picking something on a flexible protein that's more distal to the membrane, there could be a lot of flux in there. So there's just a lot of relationships that that are important as we think about targeting complex structures of proteins on the surface of the cell. And to me, that my description of receptor bio biophysics variables wasn't an intention to say I have all the answers. It was to say, look, these things are really important. And we need to spend more time thinking about these things because they will help us design the molecules of the future.

Andrew Bradbury 41:08

Do you have a publication on the SYNTHBODY concept? And if not, are you planning to publish that?

Daniel Chen 41:13

It's a great question. Of course, we will ultimately publish. I don't think we're ready to commit to when we will publish the first work from Synthetic Design Lab. Obviously, it's been an amazing team pulling through and solving problems to get to where we are today. So there's just so much data to put out there. We will need to figure out when to have such a publication. But as you know, we just presented at AACR last week. And so there is, you know, a nice amount of the initial data on the approach and where we've gotten to in that presentation. But absolutely, I think a publication on this is really important because obviously there are a lot of details that you just can't cover completely in just an oral presentation.

Andrew Bradbury 42:04

Yeah, I mean, I mean when we found a specific, that was a big discussion between myself and the CEO. He didn't want to publish everything anything. And I wanted to publish everything. And in the end, that would that would we we we published so that would and I I personally I think publication is the best PR you can get. I mean, obviously in CRO type business, you need PR, less so in in the sort of business that you're in. But certainly I think it helps to have publications out there.

Daniel Chen 42:29

Look, I fully agree. Look, I I obviously cut my teeth at Genentech. I still hold most of all well, no, almost all my values, my core values in this field come from my time at Genentech. And as you know, at Genentech, one of the things we believed a lot in was publication. And it's part of our commitment to the field, right? This is not we're in it by ourselves field. This is a community, a scientific and hopefully engineering community. And we the best way to better health outcomes is a scientific and engineering community that can work together to achieve, you know, the ultimate goals here. And it's one of the things that, you know, it's part of why I wanted to form synthetic design lab. I felt like you needed a biotech driven to do this, to push the boundaries into a future that is more powerful and was ready to embrace the complexity that comes with it.

Andrew Bradbury 43:30

So you don't think this could have been done in Genentech?

Daniel Chen 43:33

You'll have to ask Genentech that question. But I think, look, the difference between pharma and biotech, pharma can do so much more, right? So a pharma like Genentech or many others have huge teams of experts that can accomplish amazing things. But in such an organization, you have constant competition between the amazing things that you can achieve. And I think it's one of the things that is so different in biotech. We spend a lot less time arguing why someone should or shouldn't do someone something because we feel like in the time it takes to argue about it, you can essentially already have data to tell you about whether it's the right or wrong thing to do. So that's one of the obvious superpowers of biotech. But the other thing that is really special about biotech is it puts a team of people dedicated to achieving an a successful outcome of something. There's no competition. This is not, oh, we can make SYNTHBODY, or why don't we just go make two plus one T cell engagers, right? Like there's none of that. It is we live or die by succeeding in achieving something. And I think that's what it takes to push something like SYNTHBODY. You can't be saying, oh, this is hard. Let's go do something else that we know works. Every time you hit a wall, and I'll tell you, we hit a lot of walls along the way, you need to be in a position that says, no, look, we're our goal is to solve this problem. And if we don't solve this problem, we're done. We're not, you know, we can all go get other jobs somewhere else. And when you have that mindset, it's very powerful.

Andrew Bradbury 45:23

I agree. I mean, I think I think it's the real excitement of being in a startup. It's it's it's an exciting, really team-driven, sort of focus-oriented way of living, really, in a way that being part of a bigger organization just isn't. And for those out there that are young and sort of contemplating their future careers, I'd say give startups a chance is really, really very exciting indeed.

Daniel Chen 45:49

Yeah, I mean, it is it's such an amazing space. Look, I treasure my time at Genentech and Roche. I think that there are so many amazing things that you can do at a company like that. And they change those types of companies change medicine, deliver better medicines every single day. But they're not built so much to do the types of things that we're talking about here. And again, I think it's it's driven by this whole, you know, it's easy to say no to something new. Why, and it's easy to describe why something won't work, because there's no data for it, so it won't work. And it's it's when you hit hard things, you need the perseverance to really battle through it.

Andrew Bradbury 46:39

Absolutely right.

Daniel Chen 46:40

It makes for a super fun, exciting, dynamic community of biotech. And look, you can see, you can just look around today at just how much innovation and advancement in our field has been driven by this type of biotech.

Andrew Bradbury 46:54

Right. Okay. Well, I think that's a good note to end on, unless, Dan, you have anything else to say.

Daniel Chen 47:00

Well, Andrew, I always enjoy getting a chance to talk to you. It's been fun interacting at PEGS and Beyond over the years. And, you know, look, I look forward to what we can all do together as a field. So I hope it's something that we get to continue to work on together.

Andrew Bradbury 47:19

Right. Okay. Thanks very much.