Model. Predict. Discover. with Dr. Jens Carlsson | Dr. GPCR Podcast #175

[Music] Hello listeners, Yamina here. Welcome to

the Dr. GPCR podcast. Today's guest is a scientist who didn't plan to be one. Dr.

Yens Carlson grew up in a small village in Sweden where he had never even met a PhD holder. Now he leads a powerhouse

PhD holder. Now he leads a powerhouse team at UPS University combining computational chemistry and medicinal chemistry to decode how GBCRs recognize,

respond, and reshape biology. But the

journey wasn't linear. From a lousy experimentalist to a global expert in receptor modeling, Yens's story is a masterclass in chasing curiosity and

turning failure into focus. will unpack

the difference between explaining and predicting and what happens when models meet molecules and how a cold email led him straight into the heart of GPCR discovery.

Terry's Corner gives you the framework top drug hunters rely on distilled into bite-siz lessons by Dr. Terry Kakin himself. Each session translates complex

himself. Each session translates complex receptor pharmacology into action ready insight. Training drops every week, plus

insight. Training drops every week, plus live Q&As's every month. Explore it at terkinakin.com.

terkinakin.com.

This month, we are excited to spotlight Revit's pH sense reagents in the Dr. GPCR ecosystem. It's a breakthrough

GPCR ecosystem. It's a breakthrough toolkit designed to make GPCR internalization studies simpler, faster, and more scalable. With four flexible

formats, no wash protocols, and TRF detection, PHS sense gives researchers the ability to monitor internalization in real time, even at endogenous

expression levels without imaging. And a

quick thank you to our collaborators at Celtaris Research. Their fluorescent

Celtaris Research. Their fluorescent lians and real-time binding assays are enabling sharper GPCR discovery through chemical biology. All the links that

chemical biology. All the links that I've mentioned are in the notes of the show. And now, let's dive into this

show. And now, let's dive into this episode. Hello everyone. This is Yamina

episode. Hello everyone. This is Yamina from Dr. GPCR and I'm very excited to be back in 2025 recording new podcast episodes just for you this very early

morning uh in January. I have the pleasure of having with us Dr. Yans Carlson from from the University of Yans. Welcome to the podcast. I'm very

Yans. Welcome to the podcast. I'm very

excited to have you on.

Excited to be here.

Let's start at the beginning because this podcast has been in the works for a long time. I do remember getting to know

long time. I do remember getting to know a little bit more about your research and making a mental note of oh we have to make send an invitation to the podcast that took a while but then I

think scheduling this podcast also took a while better late than never let's start at the beginning could you please introduce yourself tell us who you are and what is it that you do

all right so uh I'm a professor of computational biochemistry at university since about two years pack.

Um, my group always starts from the structure. So, we're interesting the

structure. So, we're interesting the structure of the receptor and see what we can learn from the structure. And

often it's often about interactions with lians and how they modulate receptor function.

We use lots of different techniques such as molecular docking, molecular dynamic simulations, increasingly machine learning with protein structure

prediction and trying to see how we can connect pharmarmacology uh to the actual structures uh with simulations. Uh since a few years back,

simulations. Uh since a few years back, I also have a medicinal chemist. So we

make compounds that we design um and we try to answer questions such as what is the activation mechanisms?

Why do the receptors recognize this type of compound and not another? And try to design molecules with new properties that can be either chemical probes to

understand function or storing points of a drug discovery project.

Now I don't hear you actually.

Of course not because I muted myself while I was drinking a sip of my coffee.

So I love it. Apologies. It's still

early morning here. Um my question was more around did you always know you wanted to be a scientist and what led you to where you are today?

No, I did not know that. Um and I got into GPCs very late. So in 1998 I I moved from a from a small village in the

south of Sweden uh to Obsola uh to study to become an engineer. And at at that point I'm trying to remember but I don't think I knew what a PhD was like. I I'd

never met someone who had a PhD and scientists were you know in in movies or comic books. Uh so I didn't really think

comic books. Uh so I didn't really think much about that.

uh it was around the boom of biotech. So

I'd read about the sheep dolly and engineered tomatoes. So I thought, okay,

engineered tomatoes. So I thought, okay, this is the future. So I went to a program called biotechnology engineering. They needed engineers that

engineering. They needed engineers that could work with biology that hadn't really existed before.

Mhm.

Uh and during that time, um I did some internships to try to understand what I wanted to do. and I realized that I'm a lousy experimentalist.

So I did a summer internship and um I'm purifying proteins and after the internship I wanted a recommendation letter and the professor wrote it down

and sent it to me and it said didn't say almost anything about the the actual protein purification. Instead it said I

protein purification. Instead it said I was a gifted in the modeling part and I thought did I do modeling

I I I I I in my spare time I was sitting looking at structures that they had solved. Uh but what I think it meant was

solved. Uh but what I think it meant was that I was a lousy experimentalist but I read it as if I was a gifted model. So I

thought this is my thing so let's continue along that path. And at the end of uh my uh masters, I I asked the

professor who was doing molecular modeling uh to give me three names of uh professors in California because I wanted to do my thesis uh in California.

And I still have that note somewhere.

And two of those actually got the Nobel Prize about 10 years later. Uh and I emailed them and one of them replied.

Um, I ended up at scripts in San Diego where I studied how proteins u reacted to pH and how to treat water in simulations.

There was still nothing uh about GPCRs but I realized there that the people working there they were really cool like I remember sitting in the coffee room they were talking about all these

different simulations and I felt I felt at home there. So I went back to Sweden uh started my PhD uh which was then on

small molecule design like how could we design ligans of proteins as there were no structures of GPCRs uh I never en encountered them uh I

didn't know about them anything actually about receptors uh but uh working with the modeling and also I felt that that collaborating with

the experimentalist was kind of like a waste of time I saw that uh the students next to me who were protein crystalallographers they were just purifying proteins over and over again

and failed over and over again whereas I as a model I would just went to the PTB and I could get the beautiful structures look at them use them uh without having

to go through that experimental pain uh and that has changed a lot now but back then I was really focused on the computation then at the end I realized somehow that that there's difference

between to predict and explain simulations. People working on

simulations. People working on simulations spend a lot of time explaining like we get this structure, it has this function or it binds this compound and they try to explain that

but fairly few of them were actually predicting things and uh let's say you could explain what the weather was like today. Uh could that model explain what

today. Uh could that model explain what the weather would be like tomorrow? And

I felt many of those would not would fail. they would not be able to predict

fail. they would not be able to predict the weather tomorrow. They just now they didn't make things up. It was a it was a theory could be right, could be wrong, but I needed to to approach the

experiments.

uh so I started looking for a posttock uh position uh and uh if I added all the constraints I had uh uh that okay it

needs to be someone modeler that can do experiments uh and the system has to be so simple that I would believe the computations will actually work and when I added all

these constraints there was only one name uh in the world that I could go to and that was Brian Shik at UCSF Okay. So I emailed him and and I've seen

Okay. So I emailed him and and I've seen he's been on the podcast so and works on GPS source but at that time he didn't so that's not the reason why I went there

but uh he had the setup that I thought that I needed to continue. So I went there and very shortly after I joined

uh his lab he came to me and asked like hey maybe maybe you should work on GPCRs and I thought what's a GPCR

so I went back to to Google GPCR and I found okay they seem to be important protein and uh why not

there seems to be a lot of excitement around this and a colleague of mine was working on the beta 2 receptor which had then just been determined. Everyone was

so excited. So I thought okay let's try GPCRs and since that day uh it's been my main topic. So it was all just

main topic. So it was all just coincidence uh of some kind. So the

first collaborator I had there in Brian's group was Ken Jacobson who you also uh recently interviewed. uh and uh we worked on the atadenosin receptor and

it wasn't difficult to uh select targets at that time because there were only uh three structures or four maybe rodopsin

beta 1 beta 2 and generic and adenosin so let's pick number four uh and uh so I tried to identify leans using virtual

screening and that went really well uh and then I thought yeah okay so so we can predict the future uh we can't predict the outcome of experiments. So I then jumped to the

experiments. So I then jumped to the dopamine receptors together with Brian Ross and Brian Shyen and there we actually tried to predict the structure of the receptor not use crystal

structures and later crystal structures were published and they look the same as our models. I thought hey yeah it works

our models. I thought hey yeah it works modeling can work uh and that's how everything started. I love it. And and I

everything started. I love it. And and I think the the pairing with Ken Jacobson uh was was a good I know that his lab from our interview together was very

much focused on small molecules. So they

have this this plethora of molecules that you can then use as a starting point to get a sense as to how do those molecules that are known for which we have experimental data potentially

interact with it in its in receptor and use that knowledge to then predict the future there great mentor for me and he's very open-minded he actually does a lot early

modeling himself uh and uh that's really unique about the GPCR field that there's a close connection between experimentalist and models. I think

that's why I felt when I was a PhD student that that there wasn't really a good um a good collaboration but in the GPCR field there's so uh so much

opportunity for good collaboration.

Yeah. Yeah. And I think to be honest, I've always been in the GPCR field and I always felt at home bec

are very collaborative. It's very easy to get in touch with others and then you talk to one person and said, "Oh, wait a minute. You should talk to that person."

minute. You should talk to that person."

And then that's how beautiful collaborations start off and navigate. I

remember I was on the email when Brian Shy contacted Ken Jacobs and was it was like two lines. It said I heard the structure was out. Can we send you some

lians liance to test and repl yes negotiation or or about anything? She's

like yeah wow that's really exciting.

That is fascinating. I I really love it.

uh you mentioned in the beginning that you wanted to move to California.

What was the let's say the motivation there in a in a sense you restricted your ge geography but at the same time it's California and there's fantastic

universities and research labs there I think I mean it was more that I was looking for the first time I mean I went there masters then it was the weather so

I wanted to go to California uh and I had no idea then that that all these great scientists were working there But in the second case, it it I

had decided on the lab I wanted to go to. I wouldn't really have cared about

to. I wouldn't really have cared about what the weather was like. But but I think my family appreciated. I had one daughter with me and two when I left. So

we really loved living in California.

Lovely. Yeah, I was asking about the first time. Second time it was obviously

first time. Second time it was obviously with Brian Shriet's lab and the team.

Phenomenal. All right. So when when you were completing your posttock when did you or did you ever think about not going into academia maybe going into

industry and using the skill that you had developed or was it a clearcut this is I'm finishing my postto and I'm looking for faculty positions.

Yeah. So I had I knew that I was going back to Sweden because uh for family reasons and um I also knew that I wanted

to pursue an academic career and I told myself that okay I'm going to run uh for as long as it I can I have funding. I

don't want to be a person you know you run out of funding and you just struggle. So for as long as I can I

struggle. So for as long as I can I enjoy it and and I have resource to do what I want to do, I will continue and

and I've been lucky uh that that has been the case. But I still even now uh sometimes people ask you if if you're ever considered industry position. I

still think today I haven't made up my mind. If I don't find um comfortable

mind. If I don't find um comfortable myself comfortable anymore in academia, I could go to industry. No, it would be fun. Uh absolutely and I the reason I

fun. Uh absolutely and I the reason I was asking this also is because we talked previously I think it was la late last year in a different context and uh

I know that you have a consulting company or you have a consulting arm of your lab. Maybe if you can talk to us a

your lab. Maybe if you can talk to us a little bit about how that came to be and um where do you think that's going to go?

So I started this consultancy company because I repeatedly got requests uh by

email uh from either individuals or companies who wanted advice uh or actually actually work to be carried out

and I I didn't know what what to answer like how how would I be able to do that.

So then um I started after getting one of those requests again I started the company uh and that allowed me then to

be part of um advising a company that was designing legals that I was experienced in. So and I also met a

experienced in. So and I also met a there was a lot of other people uh involved that like were people that I really liked. So I I really jumped on it

really liked. So I I really jumped on it because I got the chance to hang out with all these people that I admired very much. So that was nice. Uh and now

very much. So that was nice. Uh and now uh I I have I have a few other uh things I'm doing now with with the company advising uh research groups but the

company doesn't do any actual computational chemistry. It's just

computational chemistry. It's just provide on strategy. But I hope in the future maybe we could do that. uh uh

yeah I need the academic work uh takes a lot of my time so I don't have time uh but maybe in the future absolutely the reason I was asking this is because um again thinking about the

audience and I want people to get a sense of how did you get into your current academic position but the fact that you're you're having currently an academic position doesn't mean that

you're going to retire from this academic position it's all about your preferences and where where does the research take you, but also knowing that

you have other opportunities if you'd like to pursue them.

That's very important.

All right, so let's move let's move on to the next unless you had a comment uh there to add. No, I think that I just I

just try to grab opportunities when they come to me. So during the pandemic, for example, we worked on uh enzymes

involved in source code to identify very potent compound there. Um I just love molecules like working with proteins and molecules. Uh that's my driving force.

molecules. Uh that's my driving force.

So if I see an opportunity to do that and soluble proteins are great because you can determine a lot of structures.

Uh so all proteins have their pros and cons but but still everything comes back to the receptors at the end.

Yeah. Yeah. I always I always tell people that in my mind the receptor has its own personality and it'll tell you what you need to know or if you're using

the right conditions to tell you how it likes you know to bind ligans. Is there

calcium? Is there water or no water in there? And how does that that structure

there? And how does that that structure influence the function of the protein and of the GPCR?

And they all have the same structure.

When I came back to I was in Stockholm first before I came back to Obsola. Then

there was a professor there that told me that you know these GPCRs uh the structure is known it's seven helyses

and uh uh then I realized okay maybe we have a little bit different perspective because I care where every atom is and these 800 receptors can bind so many

things using that same scaffold that's it's very fascinating to me and uh and I so I think that would keep me busy until I retire I'm I'm I'm sure I recorded

earlier this week another podcast episode and u with my guest we were talking exactly about this aspect of the fact that there's 800 receptors including the alactory ones. There is a

limited number of G-proins li even even more limited number of groin subunits yet the receptors have this amazing ability to signal and to

control so many signaling pathways and physiological outcomes and it's fascinating as you said they're all seven transmembrane domain proteins yet

they all do different things yeah I tell the undergrads that they are a GPCR like what they think what they see what they to the your GPCR basically.

Yeah. I typically introduce GPCRs to people who don't know it as GPCR. You

can see, smell, and taste thanks to GPCRs because that's the easiest way to represent it. But you're you're I like I

represent it. But you're you're I like I like your the you are a GPCR analogy.

All right. So, you mentioned at the beginning that when you started working on GPCRs, there were just a few that had structures.

Is there out of that or since then do you have a favorite one that you kind of No, I don't have a favorite GPCR and I

try really hard not to have one. Uh it's

difficult.

uh the adenosine receptor was kind of my first love you know the it was a structure there are many interest it was interesting therapeutically for

Parkinson at that time and we published several studies together with Ken Jacobson and other groups on that receptor but I couldn't I couldn't stay

there I felt that uh that you have to each GPCR has a different challenge And often there there's a structure that

triggers a question for our group like a new structure or there is a specific question that leads us to work on a specific receptor. So some receptors are

specific receptor. So some receptors are simply not suitable for answering the question and some structures you know you have to move to different targets to answer different types of questions and

and I think that really helps us also to get general conclusions that will be applicable to many different target but but that said I I I truly love going to

conferences and talk to you know the person about this receptor uh like which is like an encyclopedia of of information and knowledge that I can try

to tease out and and combine with some other knowledge to to do something new.

Lovely. Lovely. No, absolutely. I think

it's great. You mentioned in the beginning that uh I I really loved your story about the recommendation letter that said that you're a great modeler.

Uh I love the attitude as well to take it as I'm I'm a great modeler. And you

mentioned that now in your group you do also have experimentalists. Can you walk us through a little bit more about uh uh on how does how is your group divided

now and what kind of experimentalists do you typically work with in the lab?

So the project starts with a question. It

could be can we design agonists uh of this receptor or can we distinguish between agonists and antagonists?

Then we choose a receptor target and and try to connect to a pharmacology group that that could test the prediction because if we since we want to predict

the outcome of experiments the the the story has to end with an experiment. So

first we can develop a method a technique and then we want to try it.

Uh then early in the project when we purchase compounds we usually buy them from commercial libraries and as you probably know now there are billions of molecules that you can buy online from

enm databases and we can screen in the lab probably a couple of billion molecules in a couple of weeks on the super computer centers. So that's the fastest way to get going.

Then every project at some point hits the wall uh that there are no commercial compounds that are interesting anymore.

And at that point our chemist, we have one chemist and about 10 computational chemists and then he comes in and makes critical compounds that we believe must

be made but cannot be purchased.

Um I don't have any pharmacology. I

tried that a little bit inside the group but I found it difficult and uh whereas chemistry where it's it's I would say it's relatively easy to convince a

pharmacologist maybe you can try our compounds if they are experts in that receptor they're maybe running that receptor assay every week it's much more difficult to convince a chemist to make

the compounds they've been disappointed by computational chemists they've been disappointed by our prediction so many times that they often they don't even answer your emails. But

so then I realized I have to have a chemist if I want to make molecules and sometimes we want to make really uh design libraries and make very specific molecules and and then I need the

chemist but I'm not an expert in uh medicine chemistry. So I I I have a

medicine chemistry. So I I I have a colleague helping me to supervise a fantastic researcher that I have in the lab.

That's fantastic. And I think you from a pharmacology perspective, you have your um you have so many choices because there's so many labs that work on so

many receptors in general in the field.

And going back to the fact that the GPCR field in general is very generous. Um

you basically you could do just as Brian did uh you know we have do you want to test these and then Ken responding yes

type of type of situations. But I can see Brian uh sending that email. I've

not only recorded the podcast with him, but met him in person uh in 2023 at the Gordon Conference. 23. Yes, 23 in um in

Gordon Conference. 23. Yes, 23 in um in Switzerland.

And my experience there really changed how I work that I realized that to accomplish the goals I had, I had to collaborate. And sometimes it's painful

collaborate. And sometimes it's painful you know um you if you can't within international groups you often cannot get a joint grants like you have to get

funding in each country that can be difficult maybe the collaborator gets funding for a different project uh than I got um and you know the pandemic was

very challenging a lot of labs closed down so collaborating is very challenging but much more fruitful than than just focusing on on your own thing.

You learn so much about their perspectives on what is important, what is what do they actually want uh to find what type of compounds and and what why

is it important. So it's been very important for me to uh interact with the community.

Yeah. And they can tell you what your molecule does to the receptor and uh you know how how does that molecule influence the receptor and what does it

do to it and you know more information because sometimes I also work with medicinal chemists I work with computational uh chemistry people and at

the end of the day it's the biology that matters and not that the other fields don't but one makes the molecule the other one predicts that

without the biology, without testing the molecule on the receptor, you won't know, you won't learn. And having the right assay that's reproducible, that

gives you the curve and be able to look at that curve and say, "Wow, okay, so this is what's going on. This is how the data looks like." is valuable information that allows you to then go

back through that design cycle and tweak your molecules to get a sense of okay what buttons or what residues do we need to interact with on the protein in order

to create the desired pharmacology and when I talk to industry uh about my students my students who have rich industry that's one of the things uh they tell me is that you know they can

interact with the chemist and the experimentalist and interpret the data because you really need to be able to do that and coming from a mostly computational background that that takes

a lot of time actually but I I realized I I need to understand the experiments also in order to uh really use the data so we we

learn from us a little bit sometimes and and and we learn from them absolutely and I think it's also about developing that common language and be able to communicate and have productive

conversations with each other where you know the let's say the computational teams understand what the assay is. They

don't have to know how to run the assay but understand what the output is and then also from the biologist to understand what is the perspective where does the computational chemist where

does the medicinal chemist come from and how can I as a biologist provide them the information they need in order to go ahead and improve on those molecules.

Yeah. And I think a very important role is to be able to explain what your model can and cannot do. So everyone wants the answer to very complicated question, a

very difficult question and many times maybe the model cannot explain that. So

sometimes we have to just answer no that is not uh within uh the accuracy of our method like is this compound twofold better than another then I would I would

say you know nor the experiment or the computation has the resolution to tell you in many cases. So that's a most important role is is to tell you what

maybe not what can be done but what cannot be done can sometimes be more important also. I agree and and I think

important also. I agree and and I think it goes same way for the for the biologists and the experimentalists as well because acid sensitivity

uh detection limit all of these things also matter and until you have an optimized molecule optimized enough molecule you can run the experiment four

times and get a variable you know piece of data each time and then oftent times you need to be able to explain well actually it's the molecule that's not

optimized and We're in that assay wind in that you know lower assay window that doesn't allow you to very nicely capture a beautiful dose response.

Yeah, exactly. It can depend on solubility or the how the sample has been prepared and and then the students, you know, my students could say we can't trust this guy. He gets sometimes 50 microar, sometimes gets 10 microar.

Well, you know, at this stage it's the same value.

Exactly. No, I agree. I agree. And I

learned that uh learned that in kind of in the hard way because I've spent a lot of time explaining, okay, this is the assay and yes, whether whether the EMAC

is 70 or 100 with an unoptimized molecule, that's pretty much the same thing at that point. Wonderful. Wonderful. All

right, so let's go into the last segment of our conversation. And as I mentioned before we started recording is the goal with the podcast in general is to

inspire junior scientists but at the same time allow them to get a sense of what who you are and what do you do but and and last but not least get some

advice. So if you had to give advice to

advice. So if you had to give advice to junior scientists aspiring to contribute to the field what would that be?

So on a on a general uh level, I think you need to pick a problem that you can't stop thinking about. So

about. So if you're not genuinely interested in what you do and just maybe picked it because it's a hot topic or it seems

important, then you won't have the energy uh to complete the project or pursue it. uh because you will have so

pursue it. uh because you will have so much failure like we fail all the time.

Most of the time the experiments fail, the calculations fail. So you need to be able to wake up in the morning and say hey wow that problem like when I was a

PhD student I felt every time I I woke up the first thing I was thinking what what my calculations how did they end up from yesterday and I think a drive find

a problem that makes you really excited.

Uh then I think mentors is really important. Uh I've had fantastic

important. Uh I've had fantastic mentors. Uh and

mentors. Uh and you should ask a lot of questions.

Should be afraid to approach people. Ask

them what they think. Um then you find the ones that which give you uh valuable information. But don't always I I don't

information. But don't always I I don't always do what they tell me, you know, like you have to think about why are they saying something.

Yeah.

Uh and uh and then make your own decision. So when people say uh oh

decision. So when people say uh oh that's not an interesting problem or that will never work. That's something

that's like a trigger for me. You tell

me something won't work then I'll say okay we're going to work on that. And uh

so find people that that you that can be your mentor.

Um and be happy about the small successes. That would be my third advice

successes. That would be my third advice that there you need to be able to be happy about a single small experiment working. If you're only happy when you

working. If you're only happy when you get a big paper published or you get this big grant or or get promoted then uh you won't have much happiness. So we

have to be about the small things that the small things that uh excel the projects.

I love that advice. I I very much love the mentor mentoring part but also about the small successes because um I feel like these questions or these

problems are so complex. We may not have we may not be asking the right questions or we're asking the questions too soon either because we don't have the tools or and we don't have the knowledge to

answer those questions and very quickly that interest and that drive can you know dissipate. So being able to

know dissipate. So being able to celebrate small steps forward is very important to keep that interest and when you wake up in the morning you know that

okay this is the question. So, how does that work? And that motivates you to go

that work? And that motivates you to go back to the lab and run those those calculations or run those experiments. I

really love it. Thank you.

All right. Top three aha moments that shaped your trajectory as a scientist.

I think I think the moment I realized that I was useless in the lab, that was an aha. uh I already I knew that for a

an aha. uh I already I knew that for a long time but I I I had to somehow try before it was settled question uh so

that really uh set the direction for what I wanted to do. because I was very focused from that point on learning programming and molecular modeling

techniques and that's also how I met my PhD supervisor and so that was one then the second a uh

moment was during my post-docctor research finding how fruitful collaboration can be that that really changed and and that

that I had sort of the know the possibility and to or an ability to somehow interact with other people, form

projects and um carry out those uh together. Whereas I initially

together. Whereas I initially viewed science as as someone sitting alone in a room doing simulations.

That's a little bit how I I was I was raised somehow that you you worked on a problem. Maybe that's how I saw it in

problem. Maybe that's how I saw it in the cartoons. There was this crazy

the cartoons. There was this crazy scientist alone in the basement. That's

how I viewed science. But now it's much more collaborative.

Yeah. You mentioned, please go ahead. Uh

um I'll I'll comment afterwards. You

mentioned the cartoon and the crazy scientist and I had a mental image that came into mind. Please.

Uh I don't know. Um

yeah, I think I think I think those are two things. I realized collaboration

two things. I realized collaboration would be very important for me and that I'm not an experimentalist and also the what I talked about this difference

between the difference between explain and predict that okay scientists think they can explain all things but can't they predict it that's something that I

really struggle uh every time we do a project I think okay did we predict this or did we explain this I really want to do perspective

uh science that we can use these models.

We need to prove that the computational methods can guide drug discovery, not explain what happened.

Yeah. Yeah. All right. Two comments.

One, the the scientist and the comic book. And you're right. I think you

book. And you're right. I think you watch movies or you look at comic books and the scientist is sitting somewhere in a basement in the dark, hair all over

the place and it's that lonely type of work where in reality it's not at all and it cannot be it cannot be that in order to make progress

today. At least maybe there was a time

today. At least maybe there was a time where you, you know, can sit at your desk with pen and paper or a few individuals that could do that, but it's

so much more fun and moves things forward so quickly if we work together.

Agreed. Agreed. And then you said, you know, uh, predict and and explain. I

think with the the coming up of machine learning, the explain part was important in order to be able then to predict. But

now we have enough data explained to try and predict.

I think it's super exciting what's happening now uh in AI structure prediction. We have several projects uh

prediction. We have several projects uh one one published where we use alpha fold models to identify leans. So that's

uh definitely um super exciting to see both what you can do on small molecule design with AI methods and um and the structure prediction but the rule

remains the same. They have to be able to predict the weather tomorrow.

Yeah.

And we still don't know. Um I've seen some amazing things with AI prediction of uh protein or receptor peptide

complexes which alpha fold cannot have access to and um we started modeling this complex structure in the

lab and when I saw the alpha fold structure I said yeah this is it must be right. Um, I just did what I usually do,

right. Um, I just did what I usually do, you know, read the literature, find mutagenesis, but alphold, it just one second it beat me and I saw uh later the

structure came out and it was right.

Wow.

So that really shocked me. But then I then I ran into the lab and said now we have to try this out. Then it didn't work as well when I tried it on another case. But um but it's a true

case. But um but it's a true breakthrough.

Wonderful.

Learn how to do it.

Yeah. I I I think so too and I think it's also uh about how much information is there that we can start explaining what's what's the starting point in

order to then start predicting and I love the analogy about the weather because nowadays the weather is very difficult to predict almost as difficult as as predicting

GPCR structures.

Yeah. And and small for small molecules I think there is a data problem. So we

the what enabled the structure prediction breakthrough was a lot of sequences and for small molecules we don't have that information. So some new

type of data is required to really make a major advance there and we had some experiences with the new alpha 43 and yeah as expected maybe it's not working

as well as uh we had hoped.

It's a matter of time.

Yeah we'll see.

All right, last but not least, this is I think the toughest question of them all.

What should we title this episode?

I have no idea.

I typically it's it's in the questionnaire for two reasons. One, I'm

very bad at naming episodes and number two, I can't we cannot pick a title ahead of time because we don't know how the conversation will go. And so I

decided to uh involve our guests and try and find an episode title.

I think it's it's a tough one because we talked about many different things, but it was mainly I I liked very much the weather analogy and predicting versus

explaining. So I think that's that's

explaining. So I think that's that's that on on a more you know personal front on your end is really the modeler

versus the experimentalist you know or the experimentalist turned the modeler type of uh vibe that came through. What

do you think?

Uh I don't know maybe the explain versus there's a difference between explain and predict.

Yeah. Yeah. I don't know what the other titles look like. Um

well um it really depends. I've had uh and now I'm actually these episodes will have come out by then. So we have research the gene that skipped a generation computational chemistry of

allactory GPCRs. Um non-cononical path

allactory GPCRs. Um non-cononical path to a PhD via the opioid epidemic.

Um we have something around from curiosity to innovation. So it's really it does it's it's um we can be as creative as we want to

predict not explain.

Yes.

I think uh I think that's that's kind of goes through the the spirit of our conversation.

I think so too.

Wonderful. Jens, thank you so much for your time. I really appreciate we had a

your time. I really appreciate we had a I had a really fun time getting to know you more.

Well fun too. Yeah, I appreciate very much that I got this opportunity.

Thank you so much. Don't go anywhere.

Thank you to Dr. Yens Carlson for such a thoughtful conversation equal parts human and technical. From failed protein purification to shaping how we model

receptor behavior, his story reminds us that clarity comes through persistence.

Here's what you can do next. Share this

episode with a colleague. Subscribe to

the Dr. GPCR podcast and leave us a review. It helps others find the

review. It helps others find the podcast. Want to level up your GPCR

podcast. Want to level up your GPCR knowhow? Join Terry's Corner for weekly

knowhow? Join Terry's Corner for weekly pharmacology lessons and exclusive live Q&As with Dr. Kakin himself. Start

learning today at terrykin.com.

At Dr. GPCR. Our mission is to advance receptor science through education, community, and by spotlighting tools that move the field forward and Revit's

PH Sense is just that, a real solution to persistent internalization bottlenecks. With PHS sense, you can

bottlenecks. With PHS sense, you can finally track GPCR internalization in real time, even at physiological expression levels using a simple

plate-based format. The reagents are

plate-based format. The reagents are compatible with HTRF readers.

It's validated on GLP1R and MOPID receptors, compatible with all HDRF feeders. You can explore it today by

feeders. You can explore it today by visiting the link down in the description of this episode. And special

thanks to Selaris Research for supporting receptor science through innovative non-raactive ligan tools. I

also want to take a moment to thank the entire Dr. GPCR team for helping making this episode possible because receptors don't work alone and neither should you.

Until next time, stay safe.

[Music]