Latest / The BioHub - by Avetix / Episode 239 - Eran Seger - CEO & Co-Founder Protai

Transcript

Episode 239 - Eran Seger - CEO & Co-Founder Protai · The BioHub - by Avetix

• 0:06
  So welcome back to another episode of the Biohub and
• 0:09
  Avetics podcast. Today I'm joined by Erin Seager,
• 0:13
  CEO and Co founder of Prot AI. Erin has built his career at the
• 0:18
  intersection of technology, AI, and science and founded Prot AI
• 0:22
  with the ambition of using AI and proteomics to understand
• 0:25
  disease at a much deeper molecular level and ultimately
• 0:29
  improve how new medicines are discovered.
• 0:32
  In this conversation, we discussed the inspiration behind
• 0:34
  Protei, what makes its platform different, the lessons Erin has
• 0:39
  learned as a first time founder and CEO, and what it really
• 0:42
  takes to bring the world's of AI and biotech together.
• 0:46
  Pleasure to have you, Erin. How are you?
• 0:48
  Great. Thank you for having me.
• 0:50
  So it's honestly a pleasure and obviously super relevant at the
• 0:54
  moment. You know that buzzword AI.
• 0:56
  But I want to talk about your your story though first in the
• 1:01
  fact that it's pretty unique and you've how you've ended up
• 1:06
  founding A biotech company. Can you just take us through
• 1:08
  that journey and what ultimately inspired you to start Protei?
• 1:13
  Yeah. First of all, we're not a
• 1:14
  biotech or a thick bio. OK, small difference, but it's a
• 1:19
  significant difference. I think.
• 1:21
  I was always intrigued, like you said, about the intersection of
• 1:24
  technology and science. I come from a family of
• 1:28
  scientists. My dad is a cancer professor and
• 1:31
  I grew up in labs. I grew up inside the
• 1:34
  universities when I was a kid in UW in Seattle and then in the
• 1:37
  Weitzman Institute in Israel. And I was always the part of the
• 1:42
  mission statement of cure for cancer as the as the son of
• 1:45
  scientists. My wife today is also going to
• 1:48
  be a professor in RNA therapeutics.
• 1:50
  So there's interesting intersections.
• 1:54
  And I'm also a tech geek. And I was always intrigued on
• 1:57
  how we can, I can combine my passion to tech with the passion
• 2:01
  for biology and science. And following that passion.
• 2:05
  I did. My school actually started in
• 2:07
  high school in bioinformatics. It was early 2000's the Genome
• 2:11
  Project came out and I was really intrigued about this.
• 2:14
  It's exactly what I believed in. And I decided that that's what I
• 2:18
  want to do when I will be big. And my passion then was to be a
• 2:25
  bioinformatics professor. It was before AI broke, but in
• 2:30
  Israel, if you're a geek, and I obviously was a geek, you get
• 2:35
  I'm still a geek. Not only was your since there's
• 2:39
  mandatory Army Service, I found myself in the Geek Squad, which
• 2:43
  is Unit 81. It's our tech technological unit
• 2:46
  like NSA or MI 6 technology site.
• 2:51
  And I found myself as part of the founding area of all the
• 2:56
  cybersecurity space in Israel, which was very intriguing for
• 2:59
  me. And I, I built myself as a
• 3:02
  security cybersecurity entrepreneur also, but then you
• 3:08
  know, after 10 years, a very fruitful work in cybersecurity
• 3:14
  also met my wife and my Co founder to different people
• 3:18
  during the service. So it was quite fruitful.
• 3:21
  I, I decided and after also five years in the career as an
• 3:26
  executive in cybersecurity, I decided to go back and also a
• 3:32
  successful career in cyber seas. I guess tech bio was just a path
• 3:38
  I could not escape. And then I shifted to tech bio
• 3:43
  and I think the shift was the best decision I ever made.
• 3:46
  I really wake up every morning with full passion to what I'm
• 3:50
  doing. That's amazing.
• 3:52
  And and look, you're 1 of what is a lot of companies doing AI
• 3:57
  and your discovery, right? What what is genuinely different
• 4:02
  about you guys and what you're building, and what can your
• 4:04
  platform uncover that traditional approaches might
• 4:07
  miss? So I think it's not really
• 4:09
  different than other tech spaces.
• 4:12
  In the end, it's all about your data and the data.
• 4:15
  The AI is in the end only as good as the data can, the data
• 4:20
  that trains it. And we're generating special
• 4:26
  data in the structural protonmic space, while most of our
• 4:30
  competition is just retraining the same Bank of data in the PDB
• 4:35
  in different qualities, right. If you look at alpha fault
• 4:39
  bolts, Becker Lab and so on and so forth, they're they're all
• 4:43
  competing on the same infrastructure of the crystal
• 4:47
  structures. But before I elaborate about the
• 4:52
  special data we're generating, maybe I can give you a few words
• 4:55
  about what we're trying this problem we're trying to solve.
• 4:57
  So we're AI induced proximity company induced proximity
• 5:03
  basically is like a couple therapy for proteins and we help
• 5:06
  them stick together with the small molecules that stick the
• 5:10
  protein together or glue the protein together.
• 5:13
  And this is what we understood is this is an area where the
• 5:18
  traditional Co folding models usually fail.
• 5:22
  They're either not trained by the right data and they're
• 5:26
  overconfident. And since unlike single
• 5:29
  proteins, there's not enough structural data in the PDB to to
• 5:35
  become a training set for these protein complexes.
• 5:40
  And if you go to two or more proteins, they also have endless
• 5:43
  confirmations that they can fit to.
• 5:46
  So the naive AI models get it wrong.
• 5:51
  To overcome this, we're generating the data I spoke
• 5:54
  about before, and we're generating proprietary data that
• 5:58
  comes from structural protonomics, which gives us
• 6:01
  clues about the geometry of the protein compounds.
• 6:04
  So we know the distances and the confirmations between 2, two
• 6:09
  different proteins and this enables us to generate the
• 6:13
  target centric state-of-the-art model for the protein complex
• 6:17
  and this overcomes areas where the nave models fail.
• 6:22
  What this brings to the world is that it opens up the possibility
• 6:26
  for rational design and drug design for molecules that induce
• 6:31
  these connections or these glues between protein, between
• 6:36
  different proteins with molecules like loose or protein
• 6:40
  that are called molecules that are called protects that enable
• 6:43
  either degradations or other functions within protein
• 6:46
  complexes. And this unlocks a lot of
• 6:51
  targets that traditionally weren't able to, to be drugged
• 6:56
  by this approach, designed by this approach.
• 7:00
  It's interesting. So what do you believe that
• 7:03
  tells us about disease that we haven't been able to understand
• 7:07
  before? So I think it's less about the
• 7:10
  understanding of the disease, but to drug proteins that were
• 7:15
  traditionally hard to drug or undruggable because it was hard
• 7:20
  to fit the right confirmation to these complexes.
• 7:25
  And by really modeling the the ternary complex, the complex
• 7:31
  between two proteins and the molecule that they glue them
• 7:35
  together, we can change or or alter disease biology.
• 7:43
  In the end, proteins are the building blocks of life.
• 7:46
  And if we can change the way they're not only binding drugs,
• 7:51
  but the way they're functioning, we can disease the way that
• 7:58
  biology is driven and change biology of health and disease.
• 8:02
  That's the idea. Yeah.
• 8:06
  And there's also the other aspect in the fact that you are
• 8:08
  a first time founder and first time CEO.
• 8:13
  What what has been some of the biggest lessons from from
• 8:17
  starting Protei? So I think the first thing is
• 8:21
  that in the end, I come from the tech industry, like I said
• 8:25
  before, and in tech industry, you have to move fast and break
• 8:30
  things, break things along the move in order to speed.
• 8:34
  If you don't move fast and break things and spend a lot of cash,
• 8:37
  it won't work. And I think the first lesson I
• 8:41
  learned is the biology moves in a different pace in different
• 8:45
  timelines. And not like tech, you just
• 8:48
  can't take shortcuts. If you take shortcuts, the
• 8:50
  biology won't work. But on the other side, there is
• 8:54
  a venture capital clock that you need to grow exponentially to
• 8:59
  succeed. And this combination of biology
• 9:02
  and entrepreneurship is not trivial.
• 9:05
  So this is a lesson. I don't have all the answers
• 9:08
  yet, but I think that's the the big difference I had from the,
• 9:12
  you know, the other industries that they came from.
• 9:15
  But one important lesson I have learned, and this is the
• 9:20
  important specifically in in biology is to know when to cut
• 9:26
  your losses because biology is complex and you have to stop
• 9:32
  early if you're going on the wrong direction.
• 9:35
  So, but pivots basically in biology are much more expensive
• 9:40
  and it's better to start early on.
• 9:42
  And we have made several pivots since we generated the company
• 9:47
  first in indication space, but also in the way we model the
• 9:51
  biology. So it's it's very important to
• 9:54
  to stay focused and make the the right decisions fast.
• 9:58
  I, I agree with you there. I think buyers like no other,
• 10:02
  you know, where you're at the cutting edge of, of science and
• 10:05
  technology. But for it to work, you've got
• 10:08
  to be super entrepreneurial because that's the, that's the
• 10:11
  competition, you know, and you were sort of saying before we,
• 10:17
  you know, jumped on this call. Well, it's recorded
• 10:19
  mid-september, just in case I release it a little bit late,
• 10:22
  but you, you, you guys are just about to start fundraising,
• 10:25
  which is exciting in a lot of ways, but also daunting in, in
• 10:28
  others. Yeah, we were.
• 10:34
  Fortunately, we have a lot of success.
• 10:37
  And then you can say it's because of the AI, you can say
• 10:40
  it's because of luck. I guess it's a combination of
• 10:43
  both. But we have success both in our
• 10:45
  internal programs and in the collaborations we're holding
• 10:49
  with several pharma companies. And this is the anchor for our
• 10:56
  next fundraising, which is planned for later this year,
• 10:59
  early next year. So I hope the podcast comes out
• 11:02
  before we finish the race so it can be still relevant.
• 11:06
  But definitely I think that it's a hot space and we're I'm
• 11:10
  confident we'll succeed in in a good fundraising.
• 11:14
  When you're looking at potential investors as well, are you
• 11:18
  looking at tech bio, Are you looking at both or, or how?
• 11:22
  How does that work? So first of all, there's an
• 11:25
  emerging space of tech bio investors, like tech bio
• 11:28
  companies that like that are interested of this combination.
• 11:32
  Most of them, like you mentioned, come either from the
• 11:35
  tech space or from the from the BIOS space.
• 11:38
  So either it's a tech investor within a traditional bio or a
• 11:45
  bio or a bio investor within the tech investor.
• 11:51
  In the end, we're we're a hybrid company.
• 11:55
  So I think are like a good company.
• 11:58
  I think the board and the investor should be a mixture of
• 12:02
  tech and bioinvestors. That's also what we have
• 12:04
  currently in airport. We're backed by a combination of
• 12:07
  tech investors and bioinvestors and they want to continue this
• 12:11
  space. So some pharma investors, some
• 12:15
  tech investors and some bioinvestors.
• 12:18
  And they think that everyone gives you or each one gives you
• 12:21
  a different view of things, but also different set of
• 12:26
  connections and and the knowledge that is important for
• 12:31
  a company like ours. So our best hit will be a
• 12:35
  syndicate of a combination of the best tech investor with the
• 12:40
  Best Buy investor. But they have to be passionate
• 12:43
  of the combination of tech and the concept of having multiple
• 12:48
  shots and goals. So I think the, the big promise
• 12:51
  of, of including tech and generating a platform company is
• 12:55
  that not like a traditional biotech, you don't have only one
• 12:58
  drug and you take it all the way to to marketing, but you have
• 13:02
  multiple shots and goal and you have a, a wide funnel and you're
• 13:07
  also generating a, a bigger story than a traditional
• 13:13
  biotech. So this is something that in the
• 13:16
  vision, it doesn't matter if the investor comes from tech or the
• 13:19
  investor comes from bio, he needs to believe in the in the
• 13:23
  approach where you the risk management is different than
• 13:28
  traditional biotech. So then there's the there's an
• 13:32
  element of if you're talking to a tech investor, you're trying
• 13:36
  to help them understand the biology and if you're talking to
• 13:39
  a bio investor, you're trying to get them to understand the tech
• 13:42
  and and the realities of each side as well.
• 13:44
  Right, it's complex, right? I'm not sure I fully understand
• 13:47
  the complexity of this a combination and how to fit them
• 13:52
  together. But but in general, I think that
• 13:57
  the tech investors are more and more understanding biology and
• 14:02
  understanding the risks of biology and also the upside, the
• 14:07
  potential upside. So it's 10 times more riskier,
• 14:10
  but the upside is the a hundredfold, A hundredfold
• 14:14
  better than the general tech, which is usually growing
• 14:19
  linearly. But in in biotech, if you really
• 14:23
  generate the best in class drug, you can get very huge revenues
• 14:29
  in a step function very fast. So the upside is also much
• 14:32
  bigger when it comes to the economics of things.
• 14:36
  They just need to be more patient.
• 14:38
  Yeah. But that's it.
• 14:39
  Yeah, 100%. I mean, it just generally AI is
• 14:45
  becoming a big talking point in bio though, you know, even if
• 14:48
  you don't contemplate the tech side that we're talking about.
• 14:52
  Where? Where do you think the genuine
• 14:55
  impact is today and where where do you think the expectations
• 15:00
  are still ahead of what the reality is?
• 15:03
  So first, I also spoke about it in my previous answer, but I
• 15:08
  think that it's become kind of commodity even when it comes to
• 15:14
  a single protein and a single binder.
• 15:17
  And they think and also in protein nano bodies and antibody
• 15:22
  design, the AI drug discovery is becoming very good.
• 15:26
  And this is something that came from over a decade of crystal
• 15:29
  structures and alpha fold and and a very strong foundation
• 15:36
  around proteins in their static mode.
• 15:39
  But when it comes to this is the area we're trying to play at
• 15:43
  when it comes to protein complexes, they're much more
• 15:45
  dynamic and much more hard to model.
• 15:50
  And that's the next step or the next unlocking of the drug
• 15:55
  discovery. Because when you'll really
• 15:57
  understand the dynamics of the biology and how proteins
• 16:00
  interact with each other and how we can induce these
• 16:03
  interactions, you'll get you're, you're unlocking a huge area of
• 16:08
  potential drug discovery of induced proximity.
• 16:11
  And that's one area I think is very interesting to explore.
• 16:16
  Another area that's not connected directly to Proty, but
• 16:20
  they think that is, is very interesting, is not only
• 16:24
  modelling the engineering side, but the modelling of the biology
• 16:28
  itself and the cell interactions and understanding how how
• 16:35
  biology is working using AI. And these are two areas that
• 16:39
  they think 1 is more an engineering problem, 1 is more a
• 16:42
  biology problem. We decided to focus on the 1st,
• 16:45
  but maybe my next company can be on the second.
• 16:50
  I'm personally very interested about that.
• 16:53
  Yeah. And then as as you guys continue
• 16:55
  to grow, what, what are the sort of the biggest milestones that
• 16:59
  that you guys want to achieve over say the next five years?
• 17:04
  So first bring successful drugs to the clinic that's already
• 17:09
  next year, but also get them approved and show that they're
• 17:12
  much better than current, current drugs.
• 17:16
  And I think this will be the true inflection point that will
• 17:20
  make the company from a biotech to, you know, a successful
• 17:26
  revenue generating company until you're really helping people,
• 17:33
  you're not validated yet. And that's one inflection point.
• 17:37
  And the second is more on the technology itself.
• 17:40
  So with really to generate state-of-the-art models where
• 17:44
  companies that even though they raised a lot of money and
• 17:47
  there's few companies that raised billions of dollars, the
• 17:51
  last area in our space and it's becoming very, very, you know,
• 17:56
  capital intensive. I think that because they're not
• 18:00
  generating this special secret sauce data, then we can still be
• 18:06
  better even if we're we raised a hundredfold less money than they
• 18:10
  raised. And I think the secret doesn't
• 18:13
  come from only capital, but it comes from the ability to really
• 18:17
  generate the right data at scale.
• 18:20
  You guys are mainly out of Israel at the moment, right?
• 18:24
  Right, Tel Aviv, Israel. You continue to Do you want to
• 18:28
  continue to be purely Israel or will you go global or what?
• 18:32
  What does that look like? Or do you don't know?
• 18:35
  The the answer is yes. So we're, I think both for sure.
• 18:40
  We'll, we're becoming a global company today.
• 18:43
  We already have a small based in UK and a small based in US.
• 18:48
  Every comp, every area has its secret sauce.
• 18:54
  I think Israel is one of the best hubs for AI and my network
• 18:58
  is in technology and AI is very, very strong.
• 19:02
  And I think this is something we're leveraging and together
• 19:07
  was very strong academia and and knowledge base and the biology
• 19:13
  side. So the, this combination is
• 19:15
  quite unique and there's a whole pool of talent that we're
• 19:19
  leveraging that is very unique here.
• 19:24
  And when it comes to clinical development, they think that
• 19:27
  that's the area will where we are expanding to in the US and
• 19:34
  of course marketing and the the commercial stuff.
• 19:37
  And UK is the best hub in the world for medicinal chemistry.
• 19:41
  So all, all of our medicinal chemistry is currently done in
• 19:45
  UK and we're we're planning to expand that also in the UKI
• 19:48
  think in the end it will be a hub that's global and really
• 19:54
  depends on the needs. But now we're focusing, you
• 19:57
  know, to where the the team is still young, scrappy and hungry.
• 20:02
  It's a small team and we're of that moving together with the
• 20:10
  with the where, where we need to be at the single moment.
• 20:13
  So it's not the. Geography is not a limitation,
• 20:18
  Yeah. I love that.
• 20:19
  Go where the talent is. I, I, I don't know how much
• 20:21
  research you'd put in to say looking at the UK market and
• 20:25
  going that they'll leave it in medicinal chemistry or, you
• 20:29
  know, when you're looking at bringing up that you, you going,
• 20:31
  OK, well, that's what we could get in the UK.
• 20:33
  That's what we could get in the US, That's what we could get if
• 20:35
  we stay in Israel. Those kind of things.
• 20:37
  You're you're going into a lot of detail.
• 20:40
  I think that in the end that's we, we're also flexible, right?
• 20:45
  It's not the yeah, it wouldn't have to be.
• 20:48
  I think since COVID geography became flexible.
• 20:53
  Also people work from afar and also it's not really it's more
• 20:59
  about your capabilities than your physical location.
• 21:02
  And we still are from from office company.
• 21:06
  So in a specific site, we like people to come.
• 21:09
  And I think the synergy comes from people being, you know,
• 21:14
  across the corridor from each other or across the desk from
• 21:17
  each other. But at the same and the same
• 21:21
  time, I think that we have to be flexible on, on geography, on
• 21:26
  talent and and build it in a lean way like like other like
• 21:32
  more like a tech startup than like yeah, yeah, it's a.
• 21:37
  Balancing act? Definitely, Yeah.
• 21:41
  Look, that that brings me to our final question, which is a
• 21:44
  question I ask every one of our guests.
• 21:46
  Erin, I don't know if you had a chance to think about it, by the
• 21:48
  way, if you could have dinner with any scientist or innovator,
• 21:53
  past or present, who, who would it be and why?
• 21:55
  Well. There's a few that I can think
• 22:00
  of. It's a hard one.
• 22:03
  I think first of all I come from a background of physics also, so
• 22:06
  I think it it has to be a someone from the physics area
• 22:10
  and I think I'll choose Marie Curie.
• 22:13
  Amazing. First of all, because she's one
• 22:18
  of the breakthrough physical physicist in the world, I think
• 22:22
  one of the best, if not the best, but she's an inspiring
• 22:27
  person in so many ways. Her sacrifice maybe can sound
• 22:33
  corny, but I think the sacrifice and dedication are amazing and
• 22:37
  it's something that I also hope to not not sacrificing my life
• 22:42
  like she did. I don't think she knew that, but
• 22:44
  the sacrificing is part of being an entrepreneur.
• 22:48
  So I think that's a big part. But also on the scientific side,
• 22:52
  I think that her contribution in generating the X-ray technology
• 22:56
  is the basis of everything we're, you know, doing today,
• 23:01
  right? Without X-ray, there were no
• 23:03
  crystal structures of proteins and without crystal structures,
• 23:05
  no alpha fold, without alpha fault, no protei.
• 23:09
  So I think that the herb contribution is the basis of
• 23:14
  everything we're doing today in, in understanding biology across
• 23:19
  the structural biology space. And also I like Paris, so the
• 23:29
  dinner should be in Paris. Yeah, no one's ever said the
• 23:32
  location before, but look, pleasure to have you on Aaron.
• 23:38
  It's, it's been amazing and super interesting.
• 23:43
  Yeah, we'll, we'll get this out so people can hear the story as
• 23:46
  you as you guys are fundraising as well.
• 23:49
  Cool. Thank you for having me and then
• 23:51
  love your podcast. I appreciate that and please
• 23:54
  everyone like and subscribe to the Bio hub and and thank you so
• 23:57
  much again, Aaron.