YT YouTube2Text

YouTube2Text → Transcripts → Anthropic Acquires a 1-Person Therapeutic Company, ARPHA-H GIVE Program Highlight, and Much More...

Anthropic Acquires a 1-Person Therapeutic Company, ARPHA-H GIVE Program Highlight, and Much More... — Transcript

by The Biotech Voyager · 20,842 words · 3,512 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:02[music]
0:07[music]
0:12[music]
0:17[music]
0:21[music]
0:35[music]
0:42[music]
0:53[music]
0:58[music]
1:02[music]
1:09>> All right, ladies and gentlemen, welcome
1:11to the Biotech Voyager. It is
1:14uh the final show of the month of
1:15October. Thank you so much for joining
1:17us. The Biotech Voyager is the show that
1:19covers what's next in biotech, mapping
1:21the early stage signals before they are
1:23big news. We're so excited that you have
1:25joined us.
1:26Uh we've got a big lineup of shows set
1:30up for today. We're going to be uh
1:31talking with a couple different
1:32companies
1:33uh across the across the the biotech
1:35space. Obviously, a lot of news and
1:38attention on the biotech space this
1:40week. Um some pretty big signals to
1:42cover. So, excited that you've joined us
1:44uh and stick around for the interviews
1:46that will be starting around noon. So,
1:48about an hour from now. Uh got some
1:50pretty cool companies to to talk about
1:52today. Jeff, how are you doing?
1:55>> Uh we're going to find out.
1:57>> [laughter]
1:57>> I got in a fight with a lawn mower last
1:59night and apparently lost. So,
2:01>> [laughter]
2:02>> this is annoying. Also, my computer will
2:05see if I've really got to fix the
2:06computer thing.
2:08Um
2:09we're going to we're just going to go
2:10with it today, I think.
2:11>> Excellent.
2:11>> going to do it.
2:13>> Perfect. Well, I mean, before we get
2:15started, should we just take a quick
2:15look at the uh at the leaderboard here
2:18for
2:19>> Um I don't have a Do you got it pulled
2:21up? I don't have it pulled up.
2:23>> just a second here. Yep, I can add it to
2:25stage.
2:25>> I Okay. Well, yeah, so here's the thing.
2:28>> [laughter]
2:28>> Here's the thing. The platform is
2:30changing, okay? So, the leaderboard's
2:32not super up to date right now. The
2:33platform is changing, everyone, like, if
2:35you're listening,
2:38give me give me a a week more or two. Um
2:42we're making a lot of changes to the
2:44site. We're going to have personalized
2:45newsletters. We're going to have more
2:46content on the show. We're going to be
2:48doing an official Innovations in Waiting
2:51thing that I'm super excited about. Um
2:53we're going to have like
2:55characters and it's going to be a whole
2:56thing. We're trying to build the build a
2:58media brand. People people seem to like
3:00the um the leaderboard stuff, and I
3:02wanted to grow on that, so we're
3:04currently doing a lot of database and uh
3:07website like infrastructure changes. So,
3:10that's that's just I guess it's neither
3:12here nor there. It's just just what I'm
3:14doing in the background. But, um because
3:16of that, a lot of the signals are coming
3:18in a lot slower. So, um we're getting
3:21all the top ones, but like if you look
3:22at the leaderboard here, who is that top
3:24company?
3:25>> It is
3:27>> Aim Ed Bio.
3:29Korean clinical stage developer,
3:30first-in-class ADCs for cancer
3:32treatment.
3:32>> Okay, so this was this was the top
3:33company from from last time, too. They
3:36This is this has not changed. Um so,
3:39anyways, the Let's just jump into the
3:42Mythril. That's like that's like a
3:43really cool signal and a really cool
3:44name for a company.
3:45>> Right on. Yes, also shoutout Ilya, good
3:48to see you. Thanks for joining.
3:50>> Hey, I'm here, too, man. [laughter]
3:54>> Ilya, you got to say hi to Jeff, too.
3:56He's going to be sad.
3:57>> [laughter]
3:57>> All right, perfect. All right, well,
3:58let's jump into the Let's jump into the
3:59obviously the the massive signal from
4:01the last couple days that everybody has
4:03been talking about. Can I make this a a
4:04a slightly different size? There we go.
4:06All right, it's a it looks a little bit
4:07odd, but it'll work. All right, so
4:09obviously the massive signal that
4:10everybody has been talking about over
4:12the last couple days has been the
4:14acquisition of Mithril Medicines. First
4:16of all, can we just can we just take a
4:19moment to appreciate
4:21the incredible amount of companies that
4:24have been coming out recently that are
4:25based around Lord of the Rings names?
4:28>> Dude.
4:28>> [laughter]
4:29>> Yeah.
4:29>> There's so many.
4:30>> We need to like go through
4:31>> Palantir. Who There's Palantir,
4:33obviously. There's now Mithril.
4:35>> Yeah.
4:35>> What are the other ones? There's There's
4:36like three or four more.
4:37>> Bio, there's
4:39Erebor Bank. So Erebor is like the
4:41mountain that all the dwarves have the
4:43the gold inside that the dragon's
4:45hanging on to, like
4:46>> you would know
4:48the names of this obscure mountain.
4:50What's up, Ilya? I get it. I'm the
4:51second I'm the second afterthought.
4:54It's fine. We'll work on it.
4:57>> Excellent. [laughter]
4:57But anyway, yeah, just wanted to take a
4:59a chance to basically take people
5:01through this if you're not familiar. So
5:02basically, there's a bunch of cool
5:04things we need to talk about
5:05specifically with this story. First of
5:07all, this guy who has who operated his
5:11company by himself, people are quick to
5:14come in and say, "Hey, you know, he was
5:15working with some consultants. He was
5:17working with CROs."
5:18Sure, agreed. Regardless, acquired by
5:21the
5:21>> every other biotech company ever. So
5:23whatever.
5:24>> Yes, exactly. He's He's operating it by
5:26by himself. He's the He's the master
5:28coordinator. There's no one else in on
5:29the
5:30cap table, per se. Um and and so the
5:33really interesting thing here is this
5:34guy isn't a scientist by training. He
5:37has his like He He started with an with
5:39an MBA and worked as an like an analyst
5:41and has stayed mostly on like the BD
5:43side of of biotech before basically
5:47taking all of that knowledge and going
5:50off and starting Mithril Medicine. So
5:52that happened that happened a couple
5:54years ago. Um had had some successes
5:56previously.
5:58And then was able to orchestrated a
6:01large part of it using Claude and Claude
6:04workflows and whatnot and there's been
6:06lots of controversy back and forth
6:08about, you know, how much how much how
6:10much was Claude actually doing.
6:12Regardless, at the end of the day, this
6:14is such a cool thing to see
6:17and it's awesome to see that there's
6:20And I think honestly we're we're
6:21entering a new era of people realizing
6:25now what's possible. It's possible to
6:27start a company just with a couple
6:28people,
6:30operate some pretty robust workflows and
6:32start generating some some some some
6:35data that can be utilized for something
6:38like an acquisition if you if you find
6:40the right partner.
6:42>> What did he What is What is the company?
6:44Let's look at it.
6:46>> Uh Mithril Medicines.
6:47>> Yeah.
6:48>> Yeah, do you want to do you want to pull
6:49it up or or I can jump?
6:50>> Yeah, I got it here. Let's just I'll
6:51share my screen here.
6:53Let's see.
6:58>> [snorts]
6:58>> But overall, again, just an absolutely
7:01wild story.
7:02I My hope right at the end of the day
7:05here is this is something that's going
7:07to inspire the next generation of
7:08founders to be like, you don't need to
7:09have $6 million. You don't have to have
7:13a huge team of people to get started.
7:15You can just do things.
7:17>> Well, they got a pretty nice website all
7:18things considered, too. Good for him.
7:21Um okay, so grounding drug program
7:24decisions invalidated biology. One
7:26medical world model
7:28is built on valid. Okay, so this this
7:30looks like
7:31the world model. It reminds me a lot of
7:33the language that um
7:36Oh, we
7:38we had them on the show. Oh my gosh, we
7:40we announced their launch. Who was that?
7:42>> Big Picture Bio.
7:43>> Big Picture Bio. R&D cycle between
7:45experiments takes weeks. Mithril
7:46accelerates that by 32x on average
7:48delivering the audible biological ground
7:50grounded insights high stakes decisions
7:52require. Okay, so this seems like a like
7:54a data company
7:56more than anything, right?
7:58>> For sure. And on top of that, the
8:01in in terms of like the actual medicines
8:04it's going after, Um, going after two
8:06rare aggressive muscular dystrophies.
8:08Um, and the interesting thing here Jeff,
8:10I wanted to get your thoughts on cuz I I
8:12double checked this and I didn't quite
8:14realize that you could use antibodies
8:16for this. They are using a bispecific
8:19antibody that reattaches muscle fibers
8:22to the tissue that holds them in place.
8:24They're using bispecific antibodies like
8:27a staple. I thought you just used
8:29antibodies as a way to take like an a
8:31target for an immune response.
8:34Also
8:34>> Well, think about think about like
8:35T-cell engagers.
8:37>> Right.
8:37>> Right? Like they're kind of like, you
8:39know,
8:40cell adherent molecules, right?
8:43Um, okay. I'm going to have to leave the
8:45studio and come back in.
8:47>> Okay.
8:48>> Watch this.
8:49>> Regardless.
8:51It's not too bad. Um,
8:52but the anyway, folks, the really
8:54interesting thing here is okay, you're
8:56you're instead of it being a staple Oh,
8:58Jeff has left previously. He will be in
9:00in just a moment. Um, but in the
9:02meantime, we're talking about Mythril
9:03Medicines and the really interesting
9:04thing here is they're using a bispecific
9:06antibody um, as uh as a staple instead
9:11of as a instead of as a just a like an a
9:14taking an immune target for degradation
9:16or
9:17>> I would think about it more like, um
9:20I would think about it more like
9:23a cell cell adherent molecule.
9:26>> Sure.
9:26>> Now, I'm not super sure how
9:29Where did you find that? I'm not super
9:30sure how like effective that would be
9:32because antibodies they're not they're
9:34not they're they're they're they're um,
9:37not covalently bound.
9:39>> Well, that was my that was my question
9:40is like okay, what in the world's going
9:42on here? How is this possible?
9:43>> that sounds that sounds weird.
9:46>> Yeah.
9:47>> Um
9:51Yeah, I'm that's Again, where did you
9:52find that? I want to look and see where
9:53that was.
9:54Is it is it one of their programs or
9:56what?
9:57>> I'll have to I don't have it pulled up
9:58in front of me. It's just part of the
9:59post I made yesterday. Um
10:01>> Oh, was it about Mythril?
10:03>> Yeah, I made a post yesterday about
10:04Mythril.
10:04>> This is where this this image is taken
10:06from.
10:07I'm going
10:08>> Wow, it's so beautiful then.
10:10Congratulations on your
10:12on your canvas skills.
10:13>> Yes. Um but anyway, how he did it, how
10:16he was able to use Claude to some
10:18extent. So, Claude handled the
10:19operational workflow,
10:21um contracts, logistics, um experiments
10:24across the external labs. Like obviously
10:25he's managing it. Um
10:28and then he just had also like part-time
10:29consultants covering whatever Claude
10:31couldn't. So, anyway, uh Anthropic
10:33estimates that they would be able to
10:34bring this to the clinic by
10:37uh 2028 for human trials.
10:38>> Dang.
10:39>> Um and so
10:41again, I've been on this soapbox for
10:43ages. The tech bros are coming for
10:45biotech. The the the person who
10:47understands the the the operations and
10:50the the business side of this and then
10:53utilizes a an effective setup has the
10:56ability to kind of start a company and
10:59do a lot of things that are pretty cool
11:01from first principles.
11:02So,
11:04I know I wouldn't get too excited yet
11:06until until we see some positive
11:09validation, I wouldn't get It's cool.
11:12We'll see. We'll see if it works.
11:14Um
11:15that's all that I have to say.
11:16>> Fair enough.
11:17>> It's all that I got. Well, being being a
11:19lab bro myself, it's kind of like
11:22takes me took me years to do this.
11:25>> Well folks, if you have any other uh
11:27opinions, feel free to throw them in the
11:28comments and we can react to that as
11:29well. Um but uh we'll move on from this
11:32story. Obviously, this is one of the big
11:34ones of the day, but the other one
11:35that's like
11:37it's been a wild week for biotech. We've
11:38seen this. Another one that just
11:40happened that we're going to talk about
11:41next here is uh what happened uh over at
11:44the FDA. So, the FDA has uh loosened
11:48gene therapy construct rules and I
11:49wanted to take a a a moment to go
11:51through this. Um and so basically what
11:53this is built off of Um can you hear me
11:55okay, Jeff?
11:56>> You sound beautiful, Ben.
11:57>> Okay. Excellent. Um what this is built
11:59off of is the success of baby KJ from
12:03last year. Right? Uh it was a huge
12:05event. Basically, you had this uh this
12:07little newborn baby born with a genetic
12:09defect. It was going to be pretty awful
12:11unless it got fixed right away. They
12:13were able to basically deliver a
12:14personalized mRNA gene therapy to the
12:18baby,
12:19cured the baby. It's amazing, wonderful
12:21story. The The problem and the challenge
12:24has been that was a personalized gene
12:26therapy for baby KJ. It's not something
12:30that would necessarily benefit someone
12:31else who's in the same disease
12:33indication because it's specific to her
12:35mutation or I think it's a girl.
12:38It was specific to the mutation of that
12:39baby. Um and so
12:42given the current way that the FDA has
12:44been set up, that is a challenge. That
12:47means that you're not going to be able
12:49to
12:50take that exact same construct and use
12:52it for someone else. You'd have to
12:54basically start from scratch since
12:56you're making a slightly different
12:58product. This
13:00This This breakthrough This This This
13:02change in the the rules basically allows
13:04you to identify one particular disease
13:08and once you've identified that
13:09particular disease and you've designed
13:10like a construct for that particular
13:13disease, you can modify that construct a
13:16little bit without starting from
13:18scratch. You can basically start halfway
13:20through, which is a necessary move in
13:22the right direction.
13:24>> Yeah.
13:25>> Mhm.
13:26>> Yeah, which this is a long time coming.
13:27Let's be honest.
13:28>> It has been. It It It's It's taken way
13:30too long. Um but there is basically
13:32there's a really nice post from Alan
13:33Reiner. I believe he's the the CEO of
13:35Prime Medicine. Um and and and I just
13:38wanted to read through this cuz this was
13:39super interesting. FDA publishes draft
13:41guidance yesterday that could change how
13:42genetic medicines reach rare disease
13:44patients. The problem it aims to solve,
13:45a single genetic disease is often caused
13:48by many different mutations in the same
13:50gene. Until now, a therapy tailored to
13:52each mutation risked needing its own
13:54approval process. For diseases where
13:56mutation might affect a handful of
13:58people, patients and families continue
14:00to wait. This new framework would enable
14:02a company to get one approval for a
14:04disease with the boundaries of what can
14:07be customized defined up front. Once
14:11approved, the version of uh for the next
14:13patient's mutations could be added as a
14:15supplement if supported by in vitro
14:17data, evidence in cells that the new
14:20version edits the target efficiently and
14:22stays within the off-target limits set
14:25during the original clinical program. In
14:27most cases, no new clinical data would
14:30be required. Again, guys, this is just a
14:33massive massive
14:36uh thing that's been needed inside of
14:37the industry for a long time. Um
14:39this could be a significant unlock for
14:42gene therapies going forward. Abso-
14:45flippin' lutely.
14:47>> Mhm.
14:48Yeah, I mean, you There's not much more
14:50to add there, to be honest. It's pretty
14:51much I mean, you'd expect based on the
14:53biology that this would be a thing, and
14:55it's good to see the FDA kind of getting
14:56on board.
14:57>> Yeah. Yeah. Well, I I'd be curious to
14:59get your thoughts, Jeff, like
15:01um
15:02would would this like
15:04talk me What
15:06uh just like chat with me for a second
15:08with regards to how this could
15:09potentially be used in the context of
15:13like funding, right? You You do a lot of
15:15work like looking at these different
15:16funding events.
15:18Um
15:20I would say over the last couple years,
15:22right, gene therapy has seen a fairly
15:24significant slump in funding um compared
15:27to like other
15:28>> bad as cell therapy.
15:29>> Not as bad as cell therapy.
15:30>> Not as bad as oncology.
15:32>> Sure. Um but this is just such a welcome
15:36change, and I wonder if this could also
15:37have implications into ex vivo modified
15:40cell therapies as well.
15:42>> I wouldn't go that far. I don't I don't
15:44know the the analog that would be the
15:47kit that would be like in cell therapy,
15:48but I think that ultimately, when it
15:50comes to the investor framework, this
15:51just ex- expands the market the market
15:54size. Right? You're no longer going
15:56after seven patients. It's There's these
15:58There's these bigger bins, essentially.
16:00So, um and it's easier to It's easier to
16:03spread out of the initial bin.
16:05Um
16:07So, yeah, it'll be easier for
16:09This is definitely like it's one of
16:10those things where we know it
16:15I'm a capitalist, okay? I am I love
16:16America. However,
16:18there are some issues with the system.
16:20One of those issues being if it doesn't
16:22make financial sense, it's basically a
16:24non-starter.
16:25>> Right.
16:26>> Right? Like you need to have a lot of
16:27financial incentive to do anything in
16:30this kind of society. And when it comes
16:31to like creating a super expensive gene
16:33therapy for five people, it's like it
16:35would be really a really good moral win,
16:38but it's not going to make sense
16:39financially, so it's just not going to
16:40happen.
16:41Right? Uh but this can start changing
16:44things.
16:44Forgive my stupid computer for
16:47completely failing on me.
16:49>> [snorts]
16:50>> All right, perfect. Yeah, I've I'm I'm
16:53I'm seeing something
16:54I I do wonder though like to some extent
16:57does this also reduce the the the
17:00requirements for rare disease uh
17:03families and organizations to come in
17:05and have a much better shot at
17:08developing a therapy that can benefit
17:10all of the people inside of the rare
17:11disease community space, right? There's
17:13so many different
17:15like disparate organizations that all
17:18have to go after their specific mutation
17:20or their particular type of a given
17:22disease, and this might be kind of like
17:24a unifying factor for a bunch of
17:26different groups that are all in the
17:28same disease space instead of trying to
17:30go after their own thing individually,
17:32they can really unite um because there's
17:35there is now this kind of research this
17:37kind of or rather these new uh
17:39guidelines that have come out.
17:41>> Yeah. So, Ilya has got a question here
17:44um that's that is pretty interesting.
17:45I'm actually not super sure of like the
17:47actual numbers on this, but he says that
17:50often genetic diseases can cause
17:52different mutations in the same gene
17:53resulting in the same pathology, right?
17:55So, gene A gets mutated at spot one,
17:57spot two, spot three, and that causes uh
18:00you know, loss of the protein or
18:02whatever. Um how can we create new gene
18:04therapies that can target multiple
18:06mutations simultaneously at the same
18:08target in one gene? This would basically
18:10just be like gene replacement therapy,
18:12right? The question really is like how
18:14often do genetic like like monogenic
18:17diseases, how often are those caused by
18:20different mutations in the same gene?
18:23I'm not super familiar with like
18:24Duchenne's for example, but if you look
18:26at the at the at the um
18:28the the dystrophin gene, I'm not super
18:31sure where the mutations are, but I
18:33would imagine that there's probably a
18:35hot spot.
18:36Right? Cuz you need to if you're going
18:37to have truncation of the protein, which
18:39is what happens in Duchenne's, the
18:40dystrophin gene has to get truncated and
18:42you have to introduce that new
18:45stop codon there.
18:46Right? That's a single mutation. Nothing
18:49else is going to do it, right? And you
18:50just you just provide, you know, a fully
18:53functional like truncated version of the
18:55dystrophin gene that will work, um and
18:57that can pretty much just cover any
18:59mutation in the gene that causes a
19:00truncation.
19:01Right? So, that's typically how it's
19:03done, um but then there are other like
19:06more advanced methods, which would be
19:08things like CRISPR-based editing and
19:10things. I It's It gets really
19:12complicated. The best way to do it
19:14though would be just a simple gene
19:16replacement, which is what gene therapy
19:18basically is.
19:20Nice.
19:21Yeah.
19:22>> Yeah. No, it's uh it's Yeah, and I guess
19:25like the question would be for Angelman
19:27syndrome as well. Um
19:29uh I
19:30I need to I would need to go look into
19:31Angelman as as well, but I think that is
19:33a challenge, right? There are different
19:35different mutations inside of the same
19:37gene.
19:38Um
19:39>> Yeah, so Angelman syndrome is a regional
19:41uh neurodevelopmental disorder that that
19:43affects roughly 1 in 15,000 live births.
19:45Uh it's a caused by the loss of function
19:47of the UBE3A
19:49gene, UBE3A,
19:50on the maternal chromosome 15. Uh and
19:53basically the gene the gene therapy
19:55approaches would just be gene
19:56replacement.
19:57>> Yeah.
19:58>> So, regardless of where the mutation
20:00occurs, you can just replace the the
20:01fully functional gene.
20:03But, yeah.
20:05>> Fair enough.
20:07Cool. Excellent. Well, yeah. Uh
20:10again, very exciting time to be in
20:11biotech. I saw also this morning uh I
20:14have to go see if I could find it, but
20:15there was uh there's a guy I follow on
20:17on X that um was talking about how
20:19biotech is kind of poised for
20:22takeoff, and uh of all people Elon Musk
20:25reposted it and said like yeah there's
20:27so much that's happening inside of like
20:28the tech space where you know early
20:30early exponential change doesn't seem
20:32like a whole lot and then it just kind
20:34of goes vertical. So I think there are a
20:36lot of there's a lot of exciting moves.
20:38>> who is here to cover all of it? The
20:40Biotech Voyager. Tune in, baby.
20:43>> Mapping the early stage biotech before
20:45it's big news. Yes, absolutely.
20:47All right, excellent. Well, just before
20:49we move on here, Jeff, let me tell you
20:52about LabThred. Less admin, more
20:53science, and easier life. Too good to be
20:55true? Not with LabThred's integrated
20:57molecular biology project management ELN
20:59and sample management software. Let's be
21:01honest, folks, your science is
21:03complicated enough. The record keeping
21:04doesn't need to be. Enter code Voyager 3
21:07to get a 3-month discount at
21:09labthred.com.
21:11That's labthred.com.
21:14All right, Jeff, I have occupied enough
21:16time here with the first two uh
21:19>> Ben, we're going to skip we're going to
21:21skip Starton. We're going to go right to
21:22UCSF.
21:23>> Are we? Okay. Perfect. Let's do it.
21:26>> Okay, so um
21:27>> Do you want to come back to Starton
21:28after?
21:29>> We can yeah, we can come back later.
21:30>> Okay.
21:31>> So the reason that I want to skip ahead
21:32here is because it's super cool and I'm
21:33excited. So I'm going to share an image
21:35file here and this is going to be a part
21:37of the segment that we're going to be
21:38calling innovations in waiting. All
21:40right? So
21:42>> [clears throat]
21:44>> um what what we do here uh and for those
21:47of you who aren't aware, just real quick
21:491-minute overview, uh tech transfer in
21:51university, super big opportunities.
21:53University uh academic institutions get
21:55a lot of non-diluted funding through
21:57grants. They can do a lot of really
21:58interesting like highly experimental
22:00research and discover really new novel
22:03innovations and then we need those
22:05technologies to be taken out of academia
22:07into industry either by a biotech
22:09company, a big pharma company, and then
22:11those companies can get like venture
22:13capital or some other kinds of maybe
22:16public funding uh to push those things
22:18forward into the clinic. That That
22:21transition is called technology
22:22transfer, right? And you can go to
22:24technology transfer offices or TTOs at
22:28any university in in America and you can
22:30find innovations that are just sitting
22:32there looking to get licensed. And
22:34they're often times hard to find and
22:36overlooked. We covering early stage
22:38biotech, you can't get much earlier than
22:40this. It isn't even a biotech company
22:42yet, right? We can do it. We have the
22:44chops. We're going to do it. So, we've
22:46discovered over 200 of these things from
22:48places like Yale, UCSF, MIT, Harvard, um
22:52all the UC system, UWash, all kinds of
22:55places and we're going to go through
22:56them. Once a show
22:59and if you like it, we can do more. But
23:00this one is from UCSF. University of
23:04California, San Francisco.
23:06Um and it's specifically from a lab
23:07called the Wells lab. I think it's James
23:10Wells. Uh could be wrong there. We
23:13should probably know that. Um
23:15>> [clears throat]
23:15>> let's let's just check real quick.
23:17Do do do do do. Where did it go?
23:21Um I'll find it later.
23:23What this thing is is it's a bispecific
23:26antibody drug conjugate and I can't
23:28point here, but you can see it at the
23:29top there. It's the blue and green with
23:31the red little dots on it. It's a
23:32bispecific antibody drug conjugate that
23:35binds to two things, obviously, and it's
23:37got drugs loaded onto it like an ADC.
23:39>> Right.
23:39>> The first it bind The first thing it
23:40binds to is a cell surface antigen.
23:44Like most ADCs do. However, the second
23:46thing that it binds to is the LDLR,
23:49which is low-density lipoprotein
23:50receptor. Right? Typically thought of in
23:53cholesterol.
23:54Um LDLR has a very interesting
23:57capability of being internalized and
24:00routing things to the lysosome.
24:02>> Interesting.
24:03>> Right? And then being recycled. That's
24:05what LDL does. It will bind to
24:07low-density lipoprotein, bring them in,
24:09degrade it, go back to the the the to
24:11the membrane.
24:12What the So, put a pin in that. So, LDLR
24:16really good at being internalized and
24:18recycled.
24:19There are some proteins on the surface
24:21that make for really good ADC antigens.
24:25But in order cuz they're they're highly
24:27present on the surface of cancer cells,
24:28for example. However, many of those
24:31suffer from one big problem.
24:34>> What's that big problem, Jeff?
24:36>> In order for an ADC antigen to actually
24:38work,
24:39it has to be bound by the antibody
24:42and then internalized.
24:44Right? And that is highly dependent on
24:46the internalization kinetics
24:48of the antigen. Right? So, for example,
24:51if you look at um like PD-1, for
24:54example, right? Or we'll go PD-L1. PD-L1
24:57highly expressed on cancer cells, right?
24:59Doesn't really internalize. So, if you
25:01bind an ADC to it, guess what happens?
25:04It just sits there, right? Same thing
25:06with things like HER2, EGFR, CD19.
25:10They're not good internalization, they
25:12don't have good internalization
25:14kinetics, which makes them really bad
25:15ADC targets. So, what did the Wells lab
25:18do? They're going to leverage the
25:20internalization kinetics of LDLR, which
25:23is super robust, and then just slap on
25:26the um the cancer antigen. So, what you
25:30end up getting is ADCs binding to LDLR
25:34and the cancer antigen, and then LDLR
25:36goes, "Oop, I'm going to internalize."
25:38And it rips the whole it rips the whole
25:40thing down, and you can see it in that
25:42middle picture. I can't point to it, but
25:44you can see like the little arrow from
25:45the the purple receptor goes down.
25:47>> Yeah.
25:48>> Here is it's in like a lysosome,
25:50and then it will be degraded. And when
25:52it gets degraded, the membrane-bound
25:55antigen is degraded, the ADC is
25:57degraded, releasing the drug, right? And
26:00then LDLR is recycled. So, you get you
26:02get two things. One, you get the
26:05degradation of that molecule. And for
26:07example, with PD-L1, you want that thing
26:09gone, right?
26:10>> Right.
26:11>> So, that can be a good therapeutic
26:12advantage, number one. Number two,
26:14obviously, um you're going to get the
26:15ADC to release its drug inside the cell,
26:18>> Yes.
26:19>> And then number three, LDLR will recycle
26:22itself. So, you're not going to like
26:23quench the LDLR with the with the drug.
26:27Um it would just continue to recycle go
26:28to the membrane. So, really, really
26:30cool. Um I think that they're going to
26:32be using it. Let me find this real quick
26:34here.
26:36Um okay, here it is. So yeah, the Wells
26:38Lab. UCSF's Wells Lab fuses LDL-R driven
26:41lysosomal degradation with cytotoxic
26:43payloads, um claiming picomolar potency
26:46where conventional ADCs falter on low
26:48antigen tumors.
26:49So it's basically a way to
26:52um manipulate the internalization
26:54kinetics of an ADC, which people tend to
26:56overlook. Right? If you're looking for a
26:59good antigen on cancer cell for an ADC,
27:01you typically just say like let's find
27:03one that's expressed on the cancer cells
27:05and not expressed on healthy cells.
27:07>> Right.
27:07>> But a massive, massive um metric for how
27:11good that antigen will be for an ADC is
27:14the internalization kinetics. But now we
27:16can leverage those kinetics of LDL-R. So
27:19they're going [clears throat] to be
27:20going after things solid tumors, uh
27:22EGFR-positive solid tumors,
27:24HER2-positive solid tumors, uh
27:26PD-L1-positive solid tumors, and then
27:28possibly things like autoimmunity and
27:31inflammation.
27:32Um to eliminate those those
27:36those uh
27:37inflammatory cells. Yep, James James A.
27:40Wells is the is the main PI. And there
27:43are some published papers. If you want
27:44the published papers, um I'll be happy
27:46to send them to you or we can put them
27:47in the comment section. Uh but this and
27:49I can also link the um
27:51the link
27:52>> you do could you do that right now? Um
27:54and folks, just let us know if you can
27:55see it. Uh throw one of the links for
27:57the papers in the chat and see if people
27:59can see that.
28:01>> Can I comment right from here?
28:02>> I I think you should be able to. Yeah,
28:03it'll probably just go Yeah, it'll
28:05probably just go to
28:06to YouTube.
28:07>> Excellent.
28:08>> But let's just see if that works.
28:10>> Yeah, so I'm going to comment the
28:13That is the innovation um portal URL.
28:18>> Nice.
28:19>> You can check that out.
28:21>> Perfect.
28:21>> Um
28:22yeah, and the thing too is like we don't
28:24really know who's watching,
28:26okay? But if you're at big If you're If
28:27you're at Big Pharma, if you're a
28:28biotech, if you're like an uh
28:30uh an entrepreneur looking to license
28:31innovations,
28:33dude, check this thing out. I mean,
28:34we're going to and we're going to be
28:35going through this this section more and
28:37more.
28:38Um and I I tend to love this kind of
28:40stuff because it lets me just be a huge
28:43biology nerd. This is like super early
28:45innovation. Um so so yeah, it's in the
28:47chat. I can see it in the chat there.
28:49All right, we're good to go.
28:50>> Cool.
28:50>> Yeah.
28:51>> I'll put that in there, too.
28:54Just for the context of those who might
28:56be like, "What in the heck is this?"
28:58>> Yeah.
28:59>> Excellent. All right, ladies and
29:00gentlemen, that's uh that is the uh the
29:03breakthrough in waiting for for today.
29:06Uh very much focused on the the stuff
29:09that's sitting right now in tech
29:10transfer offices that you, if you happen
29:12to uh be well connected to some capital,
29:15could go and license and turn into a
29:18breakthrough medicine. Uh but don't take
29:20that name because there's already a
29:21company called Breakthrough [laughter]
29:22Medicines. Friend of the show.
29:24>> Do a Lord of the Rings name. Just
29:25continue with the with the trend.
29:27>> We got to We got to brainstorm that at
29:29some point, Jeff. All right. No, no, no.
29:31Right now. Let's do it right now. For
29:32this breakthrough in waiting, Jeff, what
29:35would given that Mithril Medicines just
29:36got acquired by Entropic, uh what would
29:39you what would you name this company if
29:41uh we're going to
29:41>> Ben, are you putting me on or something?
29:43I'm a I'm I'm a biotech marketer, but
29:46I'm not that good.
29:47>> We can brainstorm this together.
29:48[laughter]
29:49>> What is the perfect name that mentions
29:51Lord of the Rings that would be perfect
29:53for this? I don't know.
29:55Duolingo cuz there's two of them.
29:57>> [laughter]
29:58>> Duolingo.
29:59>> Not taken. Not Lord of the Rings.
30:01Horrible name.
30:03>> [laughter]
30:04>> No, I would say um
30:06maybe maybe kind of go with like a
30:07Gollum theme, you know, right? Because
30:09you've you've
30:10uh
30:11>> Why? Cuz that's the first Lord of the
30:12Rings reference you can think of?
30:14>> No, because you're you're grabbing onto
30:15something you're you're grabbing um
30:18you're grabbing onto something that's
30:19already existing that you're going to
30:22that already knows how to get inside of
30:23the cell um
30:25and you're going to use it as the way by
30:26which you can get uh the the destructive
30:29payload inside of mortar. So, Gollum is
30:31your way in. Gollum Therapeutics. There
30:33you go.
30:33>> Sméagol Biosciences.
30:35>> Sméagol Biosciences. He knows how to He
30:37>> We're going to move on. We're moving on.
30:39>> He knows how to get into mortal.
30:41>> [laughter]
30:43>> I I think I think it works.
30:45>> Oh gosh.
30:46>> The one show that we actually have
30:47viewers we're going to just be complete
30:49dorks.
30:51>> [laughter]
30:51>> Oh man, you know
30:52>> Anyways, how was your D&D session last
30:53night, Ben?
30:56>> Oh dear. All right, excellent. All
30:58right, moving on. What's the next So
31:00Jeff, you wanted to go back to Startin
31:02in Stanford, correct?
31:03>> Uh yes, let's go back to Startin.
31:05>> Cool.
31:06>> All right.
31:07>> So
31:08>> share my screen real quick to go for it.
31:10>> Yeah. Um here we go. So um the signal
31:14itself, let's go back and make sure that
31:15we have the right story here. Um Startin
31:18Therapeutics has signed a clinical trial
31:19agreement with Stanford to test our
31:20STAR-LLD alongside an investigational
31:23CAR-T therapy in multiple myeloma
31:24pairing Startin's compound with one of
31:26the most active areas in oncology. Okay,
31:28so let's let's find out what their
31:29STAR-LLD
31:32uh platform is here.
31:34Um because I saw CAR-T CAR-T in
31:39uh multiple myeloma and I needed to
31:41cover it.
31:42>> [snorts]
31:42>> So here is their website.
31:43>> Lovely.
31:44>> We're going to go to
31:46the I saw a a
31:48a thing a link for it here.
31:50Okay.
31:51Um
31:52len- lenalidomide.
31:55Lenalidomide. All right. What is it?
31:58What What are you? I'm going to guess
32:00it's going to be like a T-cell
32:01activating adjuvant. That's got to be
32:04what it is.
32:05I see T-cell fitness.
32:09They're keeping it close to the vest, it
32:11seems.
32:12Okay. Um the inaugural application of
32:15the company's platform STAR-LLD is a
32:16controlled low-dose continuous len-
32:19lenalidomide formulation designed to
32:21replace and expand oral Revlimid.
32:25Uh STAR-LLD maintains continuous
32:28cereblon and Icaros
32:31IKZF1.
32:33Oh my gosh.
32:34>> [laughter]
32:35>> Okay, what am I doing with What am I
32:37doing here?
32:38I thought that it would be Okay, so this
32:40is this is what it is. It improves the
32:41tumor microenvironment. I wish that
32:43they'd had I'm going to reach out to
32:44this company be like, "Hey, you guys
32:45need a better website."
32:46>> [laughter]
32:48>> What's happening? Do we Can we get a
32:49figure or something?
32:50Okay.
32:51I do know that like this is significant,
32:53this continuous delivery. I'm going to
32:55guess that there's probably some kind of
32:56like um
32:58medical device
32:59involved here, like a like a patch or
33:01something.
33:03Um
33:04but it's this continuous delivery of a
33:05drug that's going to manipulate in a
33:08beneficial way the tumor
33:10microenvironment, enhancing CD8 positive
33:13uh T cell activity, reducing exhaustion,
33:16um suppressing Tregs and myeloid-derived
33:18suppressor cells. Again, I'm not super
33:21familiar with with what IK7
33:24or IKZF1
33:27uh
33:28Sometimes zinc finger transcription
33:30factors that regulate development.
33:33Okay. Yeah, this is this is getting like
33:35nitty-gritty immunology here.
33:37Um
33:38>> Regardless, it's cool.
33:40>> Yeah, I like the I thought that it was a
33:42CAR T cell therapy, to be honest. This
33:43is a little bit of disappointing.
33:46But
33:47I guess it is kind of a cool opportunity
33:48for me to talk about uh Ben, is there
33:51some kind of confusion between
33:54multiple myeloma and a tumor
33:57microenvironment?
33:58Because multiple myeloma isn't a solid
34:00tumor.
34:02>> Right?
34:02>> Right? It's like a plasma cell
34:04malignancy in the bone marrow.
34:06>> Right.
34:06>> And one of the things that I learned a
34:07few years ago that I was like surprised
34:09about is that a lot of like lymphomas
34:11and myelomas actually do have tumor
34:14microenvironments.
34:16Even though they're not really tumors,
34:17right? But the bone marrow becomes
34:19extremely immunosuppressive, which could
34:20be also bad just in general. Like you
34:23don't want the bone marrow to be
34:24immunosuppressive, right?
34:26Um so in multiple myeloma, that's where
34:28the it's plasma cells in the bone marrow
34:30that are expanding.
34:31Uh so
34:33yeah, this this looks like it's going to
34:35help with um probably BCMA CAR T cell
34:38stuff.
34:40Uh
34:40>> [snorts]
34:40>> I was hoping to have a little bit of a
34:41better of a better uh uh
34:43figure here, but we can make do.
34:46Um, but yeah.
34:48They've got They've got their their
34:49clinical trial. It's at It's at
34:51Stanford, so it's probably like uh
34:53academic
34:54clinical trial. Uh, but regardless,
34:57pretty pretty nifty. I think it's, you
34:59know, it's it's also pretty cool to see
35:01like there are companies that are making
35:04CAR-T adjuvants as like the therapeutic.
35:07>> Right.
35:08>> Right. That's pretty cool. Like their
35:10one thing is a CAR-T adjuvant.
35:13>> Yeah.
35:15And also like that'll be pretty good
35:16for, if it works, like every CAR-T
35:18company would want to have it.
35:20Yeah.
35:21>> Or indeed it's a great play. Um, when
35:23everyone else is digging for gold sell
35:24shovels.
35:25>> Yeah, I guess.
35:26There you go.
35:27>> Okay. So it So let's just go through
35:29those those lines there quickly. It It
35:30enhances CD8 activation, reduces CAR-T
35:32exhaustion, improves tumor cytotoxicity,
35:35suppresses T regs. Interesting.
35:38>> Yeah, I'm not super sure like how it's
35:41doing this. It's got to be through this
35:43transcription factor network here.
35:44Again, I don't unfortunately, I'm an
35:46immunologist, but I don't know
35:47everything. I don't know what IKZF1,
35:49whatever this thing is, does. Um
35:52>> Quote of the day.
35:53>> [laughter]
35:54>> I don't know everything. I don't know
35:56anything. Who are we kidding?
35:58>> Yeah.
35:59>> Let's be honest.
36:00>> And it was yeah.
36:02>> [laughter]
36:02>> That's all that I got.
36:04>> Nice.
36:05Okay, excellent. Ladies and gentlemen,
36:07in about 24 minutes uh from now at the
36:09top of the hour at 12:00 noon EST, we
36:12will be having a discussion
36:14uh with uh with Centillion and Pace. Uh,
36:17they are both they are uh kind of going
36:19in together on an ARPA-H give award
36:22where they are basically
36:23uh get they got a bunch of money to go
36:25after individualized uh production of
36:29their of of therapeutics. Um, again,
36:31this is hot on the heels of everything
36:34that's happening inside of the FDA space
36:37where a bunch of new rules have just
36:39been made whereby you can make these
36:41personalized therapies for a given
36:43disease. So, this is a very very very
36:47timely discussion on how to then utilize
36:51the technology stack in order to make
36:53these personalized therapies at scale.
36:55So, stick around for that conversation
36:57that we'll be having in a about 20
36:59minutes or so.
37:00So, we're we're very excited for that.
37:02If you have any questions for the
37:03Centillion team or the Pace team ahead
37:05of time or questions about ARPA-H, feel
37:07free to put them in the chat and we'll
37:08get to those once we are in conversation
37:11with them. But yeah, should be a great
37:14conversation today. All right, Jeff, the
37:16next one here is
37:18Orphan Therapeutics Orphan Therapeutic
37:21Accelerator.
37:22We just had
37:24we just had Craig Martin
37:26Jeff's
37:28Jeff's long-lost
37:29>> My uncle.
37:29>> [laughter]
37:31>> In no way actually in no way actually
37:33related.
37:34But anyway,
37:36in particular
37:37Orphan Therapeutic Accelerator, if you
37:39haven't go back and listen to the
37:41episode we did with Craig a couple weeks
37:44back. You can find it here on on
37:45YouTube. You might be able to find it on
37:47Spotify as well.
37:48But with Craig, we were able to
37:50basically go through their really
37:52interesting business model where you've
37:54got some sort of clinical asset
37:56that was performing fairly well inside
37:58of the clinic. For whatever reason, the
38:00company changes direction, it runs out
38:02of money, it gets acquired, and that
38:04asset gets deprioritized. What do you do
38:06with this really promising asset that
38:08just is no longer getting prioritized
38:10inside of your pipeline? Well, you can
38:12basically give it to Orphan Therapeutic
38:14Accelerator as a nonprofit and they will
38:16continue to develop it.
38:18And so, it's a really cool interesting
38:21model that allows promising gene therapy
38:23candidates to continue to to move
38:26through to the clinic and ideally to
38:28market and get approved and the
38:31underlying IP can be approved upon
38:34for you as the company that gives over
38:36this this this therapeutic asset as
38:37well.
38:39But anyway, hot on the heels of that
38:41conversation that we had with Craig,
38:43they have actually the big signal Jeff,
38:46I don't know if you're able to pull up
38:47the um if you're able to pull up that uh
38:52the the the press release for this.
38:53>> Yeah.
38:54>> Um but they they were able to in-license
38:58an investigational gene therapy for for
39:01for a rare disease vision
39:03disease.
39:05And so the the key I can't quite read it
39:07here, but uh um
39:10I thought I thought I had
39:11some additional Yeah.
39:12>> What are you looking for?
39:13>> Perfect. Yeah.
39:14They're going after OTC D. Yeah. A rare
39:16disorder resulting in mutation of the
39:17OTC gene, which is responsible for
39:18producing
39:20two very large words and I'm not going
39:21to attempt to do on air.
39:23An enzyme that plays a vital role in the
39:25urea cycle for removing excess nitrogen.
39:27The result is an inability
39:30>> Carbamylase? Shut up, Jeff, at your
39:32ability to pronounce uh
39:34>> Ornithine transcarbamylase. There you
39:35go.
39:36>> Yeah.
39:37Well, my dopaminergic neurons are just
39:40firing on all cylinders today.
39:42>> Yeah, great.
39:43>> So the resulting inability to adequately
39:45remove nitrogen can lead to an
39:46accumulation of toxic ammonia in the
39:48blood leading to a risk of
39:50hyperammonemic
39:52crises, coma, neurological injury, and
39:54other life-threatening complications.
39:55Okay, so this does go beyond the eye.
39:57Interesting.
39:58But anyway, basically again, they were
40:00able to take this really promising gene
40:02therapy candidate. For whatever reason,
40:03it's not getting picked up by
40:06it's not getting picked up and so they
40:08have the ability to basically take that
40:10asset and move it forward into the
40:12clinic as a nonprofit. Really really
40:15cool approach. This is something that I
40:18think we will end up seeing more and
40:19more of as well
40:21as it gets easier and easier to interact
40:24with the FDA to some extent, right? As
40:26it relates to these different gene
40:27therapies, the manufacturing
40:29technologies come online so we can get a
40:30little bit closer to a to a push button
40:33get your get your manufactured product.
40:36Um so yeah, if you aren't following
40:40Craig Martin or Orphan Therapeutics
40:42Accelerator, by all means go and do
40:43that. They're doing some pretty cool
40:45stuff.
40:47Um and this is just kind of another
40:48indication that they are actually able
40:50to take clinical stage assets and start
40:53developing them.
40:54Um, so it's it's it's an interesting
40:56it's an interesting model. You had
40:57mentioned earlier in this conversation
40:59Jeff, right? About how
41:01sadly right inside of America a lot of
41:03different times you aren't really going
41:05to see any sort of
41:08uh move forward unless it the the the
41:10money makes sense, unless there's like
41:11commercial viability here.
41:13>> Yeah.
41:13>> This is an interesting avenue by which
41:17patients can still get served and you
41:19don't have a large farmer that needs to
41:22make $3 billion off of it in order to
41:24justify going after it.
41:25>> Yeah, I still don't know enough about
41:27the financials to know how like like how
41:29does Craig pay his salary, you know?
41:32>> Yeah.
41:32>> got it they're getting money somehow.
41:34Like there's got to be money coming in.
41:36Um, and I'm wondering like is I think
41:39that it was um Terry Terry Peraulakis
41:42>> Yeah.
41:42>> who who their big thing is like getting
41:44those waivers.
41:46>> Yeah.
41:46>> Right? Like you get you get these rare
41:47disease approvals and you get a waiver
41:49and then big farmer will pay you oodles
41:51of money for that. So like that could be
41:52a financial incentive there. Like a
41:54business model is to sell to big farmer.
41:56Um
41:57>> I think that is one part of it.
41:58>> Yeah. Yeah. So
42:00>> also the other way that he can
42:01[clears throat] make money, right? Is
42:02like the companies can actually give
42:04them uh
42:05they can actually give them some
42:06investment to develop their IP
42:08portfolio, right? Because if you're
42:10developing this one clinical asset and
42:12there's a wider IP portfolio of like the
42:14capsids and the
42:15and the tran and and the constructs and
42:17everything else. Um, your entire IP
42:19portfolio benefits from that work that
42:21Craig and his team are doing.
42:22>> Precisely.
42:23>> Yeah.
42:23>> Yeah.
42:25Excelente.
42:26>> Cool.
42:27>> All right. Um, moving on here to the
42:30next story. We got about 20 minutes
42:31left. Um, um
42:33>> Can we actually go to Strand
42:34Therapeutics real quick?
42:35>> Strand Therapeutics. I don't have that
42:37>> got 20 minutes here. Yeah, I I got it
42:39here. I've got it here.
42:41>> All right, I'll make up a I'll make up a
42:43another one here just quickly.
42:45>> Pink.
42:47So,
42:48>> [clears throat]
42:49>> big news.
42:50Strand Therapeutics dosed their first
42:52patient um and I think if you were here
42:55on Tuesday, I was on like a cytokine
42:58kick, right? And I always mention this.
43:01The very first
43:03unless you want to include Coley's
43:04toxin, you nerd.
43:06Not including Coley's toxin. The very
43:08first immunotherapy for cancer was
43:13cytokines. IL-2, interferon alpha.
43:17Right. Those are the very first ones,
43:18right?
43:20However, they're very difficult to do.
43:23They're hard to to administer because
43:25they're immunostimulatory,
43:27right? So now companies are trying to
43:30put them on to oncolytic viruses, mask
43:33them with antibodies, have all kinds of
43:35weird like mutine formulations of them
43:38to make them just active in the tumor,
43:39right? Strand Therapeutics
43:42delivers it via
43:44mRNA.
43:46Right? And I'm trying to figure out
43:48here. So the big the big signal here was
43:50they they dosed their first patient with
43:52an IL-12 encoding mRNA, STX-003,
43:56um in the Ascent-1 trial. What exactly
44:00is the indication here?
44:04It's got to be tumor oil. Yeah, so it's
44:06So STX-003 builds on Strand's clinical
44:08experience with XTX-001. It's intra-
44:11intratumorally delivered
44:13self-replicating Ooh, self-replicating
44:15mRNA therapy expressing IL-12 in
44:18patients with advanced solid tumors.
44:20That's super cool.
44:22Okay, so also, I want to take this time
44:25to talk about abscopal effects.
44:28>> Abscopal
44:29>> Big word, big word, super nerdy concept,
44:32but it's actually super important with
44:33these kind of things because you could
44:34say
44:35Okay, Strand.
44:38IL-12 injection into the tumor is
44:40supposed to help with
44:43immune stimulation, okay?
44:45>> Right.
44:46>> But normally, cancer the primary tumor
44:49doesn't kill you. The met- the
44:51metastatic diseases are is what kills
44:52you, right? Like a breast tumor isn't
44:54going to kill you. Right? A skin tumor
44:56isn't going to kill you.
44:58However, if the skin tumor goes to your
44:59brain or it goes to even like a bone
45:02tumor isn't going to kill you, but it
45:04can go to the liver. It can go to the
45:05kidney. It can go to the lungs. It can
45:07go to the brain. Those will kill you.
45:09You need to be able to have a local
45:11delivery
45:12act systemically.
45:14>> Right.
45:15>> Right? This is where abscopal effects
45:17come in.
45:18A lot of oncolytic viruses rely on it
45:20and a lot of therapeutics like like
45:22Siren Biotechnology, Siren doing like an
45:25intratumoral injection of a AAV that
45:27delivers interferon beta. Right? Kind of
45:29the same thing as this.
45:30Um what will happen is the initial tumor
45:33you'll get the IL-12 activation and
45:34everything and you will get an immune
45:36response inside that tumor and then
45:38guess what? Dendritic cells, T cells, B
45:42cells, NK cells to a certain extent will
45:44travel to draining lymph nodes
45:47and prime more cells. They'll they'll
45:49they'll they'll have more activation in
45:51the lymph nodes and then you can have
45:53those cells leave the lymph node, go
45:55through the lymphatic system and end up
45:58in metastatic tumors.
46:00>> Right.
46:00>> Killing the
46:01tumors there as well. That's an abscopal
46:04effect. These these kinds of things
46:06heavily rely on abscopal effects. But
46:10then you get a really big benefit. The
46:12IL-12, which would be which would be
46:14toxic systemically,
46:16is just happening in the tumor. It's a
46:18local injection of IL-12.
46:20>> Mhm.
46:21>> Right? So it says here that
46:23circular RNA with genetic logic
46:25circuits, that could be interesting to
46:26look into, that switch on only in
46:28targeted cells or tissues.
46:30>> [laughter]
46:32>> Hang on.
46:34Okay, we're going to we're going to
46:35spend some time on that.
46:37>> Yeah.
46:37>> to spend some time on that because
46:39typically when you think of genetic
46:40logic circuits, you think of DNA.
46:42>> Right.
46:43>> Right? Companies like um Trogenics, for
46:46example, they'll have a virus with a
46:49plasma
46:50and that plasma
46:53uh promoter or tissue-specific enhancers
46:56that will only activate in certain
46:58tissues.
46:59mRNA
47:01actual regulatory step would be
47:02translation. Right? So, now we're
47:05talking about uh logic logic uh mediated
47:10translation of mRNA?
47:13I'm sorry, who are you? Strand
47:15Therapeutics? Let's look into this.
47:19Um yeah, what is your technology? Hello.
47:22Single-path LNPs, signal lock RNA. Why
47:25am I not getting anything here?
47:28Okay, this is not what I want. I want
47:29this.
47:30To ensure that protein payloads are only
47:32expressed in target cells, we've
47:32genetically programmed logic circuits
47:34consisting of microRNA sensors that
47:36control the location, timing, and
47:38intensity of protein expression. Learn
47:40more. Uh yeah, we're going to learn
47:42more. To develop signal logic RNA, we
47:43begin with our microRNA database,
47:45correlating thousands of natural micro
47:47microRNAs across hundreds of tissues
47:48like a heat map. We use proprietary
47:50methods Oh, dude, that's super cool.
47:53Okay.
47:55Okay.
47:56Dude, that's actually really
47:57interesting. So,
47:58microRNAs
48:00will degrade mRNAs.
48:02Right? And the
48:04>> the expression of microRNAs is
48:06tissue-specific.
48:09So, if you include a microRNA binding
48:12site on your mRNA,
48:14>> Ah.
48:16>> and you you can you you exclude that
48:19site that the binding site of microRNAs
48:21that are expressed in your tissue of
48:22choice, you will get degradation in
48:25everything else, and you'll get
48:26persistence of the mRNA in your tissue.
48:28>> Negative selection. Negative selection.
48:31That is so nifty.
48:33>> Really, really interesting. Yeah, so you
48:35just load it up with a bunch of microRNA
48:36binding sites that will degrade it in
48:39skin cells and this and that, and then
48:41you can have We only want it to turn on
48:42in cells that don't have these
48:44microRNAs. Or maybe there's a microRNA
48:46that that protects it from degradation,
48:48and you can have it only bind in those
48:50tissues.
48:51>> Right.
48:51>> That's really cool. So, this is a
48:53microRNA-based
48:54therapeutic, or I guess
48:56microRNA-centered
48:58mRNA therapeutic for cancer. Dude, but
49:01biology is so cool. Scientists are so
49:04smart. What are we doing?
49:06There's the cool little graphics, too.
49:08This This looks interesting.
49:10>> [laughter]
49:11>> That's actually pretty straightforward.
49:13Here it is.
49:14If target cell, then expression on, else
49:17expression equals off. And this will be
49:19the microRNA piece, like a binding site.
49:22>> Yeah.
49:23>> That's really cool.
49:26>> That is indeed a really cool graphic.
49:27>> Can we give Can we Can we Can we stop
49:29and appreciate the website for a second?
49:31In a In a sea of of of somewhat dubious
49:35uh websites that we'll look at for
49:37biotechs, this one is gorgeous.
49:38>> Yeah, especially like therapeutic
49:40companies. A lot of times it doesn't
49:42really matter, but with with with
49:43technologies like this where you have to
49:45like really explain how it's working,
49:47like it's probably beneficial to have a
49:48nice a nice website like that.
49:51>> Definitely.
49:51>> Oh, they got a video and everything.
49:53Wow.
49:53>> Can we Can we watch the video? We got
49:55time.
49:56>> Let's see.
49:57There won't be any audio. However, hold
50:00up. Let's try this real quick because
50:03I'm going to remove that and then I'm
50:04going to add it again.
50:06Stop share.
50:08Share screen.
50:09Here we go.
50:10Um
50:12I can share it with audio now, and I'm
50:14wondering if that makes a difference.
50:15Let's try it.
50:20I don't hear anything, either.
50:22>> Maybe they don't have any.
50:24>> Yeah, there may not be any audio.
50:28>> The more damage they do to healthy
50:29cells,
50:31some of our best tools have been kept
50:33out of the fight against cancer.
50:37>> But what if we could tell a cancer drug
50:39exactly which cells to attack?
50:42>> What if
50:42>> cells to avoid? See, this is super cool.
50:45>> Yeah.
50:46>> This is super cool. And the one of the
50:47biggest issues with mRNA-based
50:48therapeutics is like the systemic
50:50toxicity. But leveraging microRNA
50:52binding sites and I guess LNP-mediated
50:54delivery and stuff like that.
50:56>> Yeah.
50:57>> Yeah. So, they're pretty much like
50:58they're building a tech stack to be able
51:00to deliver mRNA
51:03to the right cells and have it expressed
51:04in the right cells.
51:06>> Yeah.
51:08>> Wow.
51:09Strand.
51:12Who is
51:15Who are the leadership at Strand?
51:17Isn't it Is this
51:18Is this one of Doudna's companies or is
51:20that I'm not thinking of the right
51:21company?
51:22>> You're probably thinking of Scribe.
51:24>> Scribe, that's what I'm thinking of.
51:26You're right.
51:28Come on.
51:33Jacob, come on the show, bro.
51:36>> Jacob, we would love to we would love to
51:38talk shop.
51:40>> He looks familiar.
51:42>> Jason Luke, he does look familiar.
51:44>> Definitely connected on LinkedIn for
51:45sure.
51:48Nice. Yeah, we'll have to to reach out
51:50to Jacob Becraft
51:53and have him come on. It's super super
51:54cool. Of course, engineers at MIT,
51:56that's obvious.
51:58Where else would it have come from?
51:59>> [laughter]
52:01>> Super cool.
52:01>> Yeah, really cool.
52:03>> Excellent. Okay, ladies and gentlemen,
52:05we're about 10 minutes away from our
52:06first interview where we will be talking
52:08with the the leadership over at Pace Bio
52:11and at Centillion.
52:14They have received a large contract from
52:16ARPA-H, my favorite government program.
52:18I can't believe I'm saying that, but
52:20that is the that is the case. As a
52:22Canadian, I'm praising a US
52:24a US a US government program.
52:27>> [laughter]
52:28>> Um and in particular how they are
52:30developing some really powerful ways to
52:33go after these personalized
52:35therapeutics, which is going to be
52:36really interesting in the context of
52:37everything else that's going on inside
52:39of the the FDA rules and whatnot. So,
52:41stay tuned for that. That'll be starting
52:42in a couple minutes. I see that that
52:44Peter has already joined the studio, the
52:45green room 8 minutes early. We love it
52:48when our guests join early. Thanks,
52:49Peter. We can see you smiling. We'll
52:51bring you on in about 8 minutes once
52:53once Harrison's joined as well. All
52:55right. So, the next one, Jeff, just
52:57before we bring them on is the next the
53:00next signal here is Aviado Bio.
53:03>> Mm.
53:03>> [clears throat]
53:04>> What do you got for us?
53:05>> All right. So, if I can share my screen
53:08here just briefly.
53:13Okay.
53:14Um all right, this is very this is very
53:16small. Okay. Anyway, Aviado Bio super
53:19cool company working inside of the gene
53:21therapy space and they are using what
53:23they're calling a V-Mix platform for
53:26AVB-406 in Alzheimer's and Parkinson's
53:29and they just announced some pretty cool
53:31data at
53:33ESGCT, excuse me.
53:36And basically what they're doing
53:37>> is ESGCT? Oh, the European Society.
53:40>> Yeah.
53:41>> They're a bunch of copycats. Which one
53:42came first?
53:43>> Probably ESGCT.
53:45>> [laughter]
53:47>> We You're right.
53:49If one region of the world is going to
53:51be copycats, it's definitely not them.
53:53>> [laughter]
53:54>> Which came first, York and England or
53:56New York?
53:59>> [laughter]
54:00>> What came first, Philadelphia and Asia
54:02Minor or Philadelphia,
54:04>> [laughter]
54:05>> Right.
54:06>> Pennsylvania?
54:08Exactly. Anyway, just wanted to take a
54:10little bit of time to go through this.
54:12They've they've basically got this
54:13really interesting MAPT mRNA knockdown
54:16situation
54:18where they have presented their first
54:19clinical data. It looks really good.
54:22They have primarily focused on
54:25on Alzheimer's and
54:28and My goodness.
54:30Alzheimer's and what was the other one?
54:32Parkinson's, sorry.
54:34Alzheimer's and Parkinson's two diseases
54:36obviously that have loads of room for a
54:38lot of improvement inside of the space.
54:40Not a whole lot of good options at this
54:41point.
54:42And so the main the main component here
54:46is their V-Mix platform uses an
54:47adeno-associated virus to deliver gene
54:49silencing instructions
54:51and is designed to enable durable
54:53reduction of disease disease-causing
54:54gene expression following a single
54:56administration.
54:57AVB-406 combines this platform with a
55:00blood-brain barrier crossing capsid and
55:03a neuron selective promoter to silence
55:05MAPT, the gene encoding tau as a
55:07potential treatment for Alzheimer's
55:09disease. So, uh yeah, I wanted to take a
55:12chance if we can to go look at their
55:14VMAX platform. I wonder if they have it
55:16up. I I hoped that I would have
55:20the ability to to look at it, but let's
55:22see here.
55:23Um
55:25their VMAX platform.
55:27>> This is RNA RNA interference.
55:30>> Yeah.
55:31It doesn't look like they've got a
55:33dedicated page for it. So, that is a
55:35shame.
55:36Um
55:37but overall, just really exciting to see
55:39Okay. Uh yeah, RNA silencing system for
55:42targeting RNA modulation for gain of
55:44function targets and diseases.
55:45Um so anyway, that's just kind of the
55:47overall, I guess maybe here uh of the of
55:50the different approaches that they're
55:51taking. But, it's really interesting to
55:52see just how they're they're uh the the
55:55progress that they've they've had. I've
55:57been following Avidity Biosciences for a
55:58while, and it's uh and it's nifty to see
56:01uh you a gene therapy uh company that's
56:03got some some hands.
56:05>> Hang on. I found it. Hang on. Hold Hold
56:07the phone. Yes.
56:08>> phone. Take it, Jeff. Take us away.
56:10>> and make sure Okay. So, this is going to
56:11be
56:13Bam. Bing bop boom. There we go. Okay.
56:16It's an AAV-delivered RNA interference
56:19platform.
56:20>> That's illegal.
56:21>> Which it
56:23>> [laughter]
56:24>> Right. So, basically, instead of just
56:26delivering the RNA interference are the
56:28RNA molecule as an RNA, you encode it in
56:32a in a in a cassette on an AAV genome
56:36and put it in that way, and then it's
56:37just continuously being expressed, which
56:39is like this durable, you know,
56:41expression of of uh
56:44of siRNAs.
56:45>> Yeah.
56:46>> Or sorry, RNAi.
56:47>> RNAi.
56:48>> RNAi.
56:49>> Yes.
56:49>> Interesting. Okay, so
56:50>> And you can do multiplex multiplex mRNA
56:53cassette microRNAs. Um so, you can have
56:55multiple multiple siRNAs being delivered
56:58via the same the same AAV. They're
57:00normally really small. Like like an
57:02siRNA, it's like, I don't know, like 15
57:05to like 30 nucleotides.
57:07>> Right.
57:08>> Or maybe it's a little bit longer, but
57:09like the actual like binding site is
57:11like 15 nucleotides. Right? So, like
57:13maybe 100 nucleotides total for the
57:14whole thing. So, you can easily throw on
57:16what?
57:18>> [snorts]
57:18>> 300?
57:19>> [laughter]
57:19>> Right.
57:20>> A bunch of them.
57:21>> They're all pretty small.
57:23>> Yeah, that's interesting.
57:24>> Uh
57:25>> Yeah.
57:25>> To go after more than one thing is just
57:27so nifty.
57:28>> Yeah.
57:29>> Excellent. Okay.
57:31Uh well, Jeff, uh that that's that's
57:33just about it here for for Avi Aviado
57:36Bio, super cool company. Um we might
57:38have an opportunity to chat with them
57:39next week to get the them to take us
57:41through the the Vmix platform in a bit
57:44more detail.
57:46Uh but ladies and gentlemen, uh both of
57:47our both of our guests uh have joined
57:50the green room. So, we're just going to
57:51bring them on. We're we're interviewing
57:53two people at the same time, which we we
57:55haven't quite done before. So, we've got
57:57a a new format.
57:58>> Burke. We we did four at once. We're
58:00good. We got this.
58:02>> Exactly. Uh so, we'll probably not do
58:04the the normal drum roll here, but we'll
58:06just bring them here here into the
58:07studio. Here's Jeff and myself, and then
58:10we'll bring in uh Harris and Peter.
58:12Gentlemen, welcome to the Biotech
58:14Voyager. We're so pleased to have you
58:16here.
58:18>> Hello. Really pleased to be here.
58:20>> Excellent.
58:21>> Thanks for having us.
58:23>> Glad to. All right. So, I I guess just
58:25before we get started
58:26>> Harris, are you are you um on the
58:28surface of the sun right now?
58:30>> [laughter]
58:32>> What's going on?
58:35>> I tried and I tried. I I really I really
58:37don't know what's
58:37>> wrong with my camera.
58:39>> [laughter]
58:40>> We'll we'll take it anyway. All right,
58:42excellent.
58:43>> Um yeah, so I guess just before we get
58:45started uh gentlemen, uh given the news
58:48that has come out over the last couple
58:49days with the FDA and the rules that
58:51they have made, um I think this is a
58:53obviously a very timely conversation,
58:55but before we get there, could you guys
58:56just quickly give
58:58uh a a quick introduction each? Uh
59:00Harris, can we start with you?
59:02>> Yes, absolutely. Hello everybody. My
59:04name is Harris Harris Makatsoris. I'm
59:06the founder of uh Centurion Biosciences
59:09Incorporated.
59:10Uh it is uh a UK-based company and uh
59:15our subsidiary in the US,
59:18which we established
59:20a couple of years ago or so. Um, um,
59:23what Synthego does, we're building the
59:26infrastructure for programmable
59:28medicine.
59:29Um, we have worked
59:33in fact just before the pandemic and and
59:35during the pandemic on
59:39our biofoundry in a box
59:41and what it does, it brings together,
59:45you know, really the full technology
59:46stack for RNA manufacture. So, it starts
59:49with our core innovation, which is a
59:51really new novel way of of making these
59:54biological materials and using automated
59:59continuous flow methods.
1:00:01These were methods that we have been
1:00:03developing with with my team for for
1:00:06some years actually well before the
1:00:07pandemic and before actually RNA was was
1:00:10a thing
1:00:11even. And integrates purification,
1:00:16the
1:00:17formulation and encapsulation steps. We
1:00:19have done a lot of work on lipid
1:00:21nanoparticles as
1:00:23of course expect, but we also tried and
1:00:26made them work on other things as well
1:00:28and um
1:00:29putting it all together with a digital
1:00:31quality management system in a modular
1:00:33GMP compliant platform. Now, we already
1:00:37have installations in the UK which is
1:00:40about to to embark further installations
1:00:43in Southeast Asia
1:00:45and North America and we're very excited
1:00:48about that.
1:00:50We are ramping up our sites in in the
1:00:53United States and
1:00:56this is also something that makes us
1:00:58extremely extremely exciting. And
1:01:02ultimately with the biofoundry in a box,
1:01:06we are delivering reproducible RNA
1:01:09production from discovery all the way
1:01:11through to to clinical supply. And what
1:01:14this does for for innovators, for
1:01:16biotechs in particular, it completely
1:01:19what completely
1:01:21it's largely the risks you know that
1:01:23that journey from from discovery through
1:01:25to a regulatory compliance and and and
1:01:29further on through the clinical trials
1:01:32using exactly the the same platform.
1:01:35And in a nutshell, this is this is what
1:01:38we do. I mean the the company combines
1:01:40both engineers and scientists. So
1:01:43sometimes
1:01:44we joke and we say Sentium suffers from
1:01:47a you know split personality disorder
1:01:49>> [laughter]
1:01:50>> bringing all these disciplines together,
1:01:52but you know it has worked really well
1:01:53so far.
1:01:54>> Excellent. All right, and then quickly
1:01:57over to you Peter and then we'll reveal
1:01:59at the end whether or not you and I are
1:02:00related.
1:02:02>> [laughter]
1:02:03>> Sounds good.
1:02:04Yeah, so I'm Peter Peter Abbingh. I'm a
1:02:06site head here at Base Life Sciences in
1:02:08Boston.
1:02:09I myself have about 25 years of
1:02:12um
1:02:14is history in in drug development from
1:02:16early discovery, platform development,
1:02:19clinical translation
1:02:21and manufacturing.
1:02:23At Base Base Life Sciences we're CDMO.
1:02:27We have a very extensive background in
1:02:29analytical um
1:02:31expertise.
1:02:32We are focusing really heavily on
1:02:34characterization, formulation
1:02:36development and manufacturing
1:02:38as well as release testing of various
1:02:39drug products and these include genetic
1:02:41medicines such as the RNA LNP modalities
1:02:46that you know is a large part of this
1:02:48gift program
1:02:49as well as other biologics and small
1:02:52molecule modalities. So quite broad
1:02:56but yeah super excited [clears throat]
1:02:57to be part of this program.
1:02:59>> Phenomenal. Excellent. All right
1:03:01gentlemen
1:03:02very excited to have you with us today.
1:03:04This is this is a this is a conversation
1:03:06I've been looking forward to especially
1:03:07just kind of in the context of
1:03:08everything that's happened over the last
1:03:09couple years. Earlier in the show we
1:03:11were talking about baby cakes right
1:03:13right and just like the the the
1:03:15watershed moment that was and then we've
1:03:17got the we've got the the new the new
1:03:20draft guidance from the FDA with regards
1:03:21to how we can kind of look at
1:03:22personalized therapies.
1:03:24So again, I think this is going to be a
1:03:26really interesting conversation given
1:03:27how you guys are are going after this
1:03:29and trying to figure out how can we how
1:03:31can we scale individualized therapies.
1:03:35So I guess maybe like just to bring
1:03:38people up to speed here.
1:03:40Can you can you can you guys just help
1:03:42us elaborate like what is what has been
1:03:44the the challenge pre- viously? Why is
1:03:46there a need here for personalized gene
1:03:48therapies to start with? Why isn't it
1:03:50that we can just you know, make one
1:03:52therapeutic gene, stick it into
1:03:53everybody and and it does the same job?
1:03:55Why is it that there is this need for a
1:03:58personalized therapy across these
1:03:59different diseases and again, maybe we
1:04:01can start with you Harris.
1:04:03>> Okay. Well, yeah, that's that's a really
1:04:05really good question and and actually is
1:04:08is at the core of
1:04:10what we have been doing in the past few
1:04:12years. So RNA is indeed considered in in
1:04:16you know, by many a platform in in a
1:04:19broader sense. Of course, this is this
1:04:20is a very
1:04:21another you know, item of of debate. But
1:04:25nevertheless, as as you mentioned Ben,
1:04:27you know, you take a gene, you stick it
1:04:29on you know, backbone or something and
1:04:32then then you you can make the medicine
1:04:34using exactly the same
1:04:35or very closely related sort of process.
1:04:39So we are seeing and
1:04:43we we do think RNA is changing medicine
1:04:47already.
1:04:48Trouble is
1:04:50manufacturing remains the same, okay?
1:04:52And it does remain a bottleneck. There
1:04:54is a lot still that has been borrowed
1:04:57for very good reasons
1:04:58by conventional you know, protein
1:05:01manufacturer. And of course, I mean
1:05:03this is this is completely
1:05:05understandable. But then then when we
1:05:07started working in that space, we
1:05:09thought, "Okay, are we are we kind of
1:05:11missing a trick here?" So so we have
1:05:13this new modality. Can we can we think
1:05:16differently and do better? Okay? In the
1:05:19sense that
1:05:20you know, we can we can make these
1:05:23materials really in in very very small
1:05:25footprints, very very rapidly. And why
1:05:27do we have to change, you know, scaling
1:05:30up from from you know, work done on the
1:05:32bench and then scale it all the way to
1:05:35to essentially, you know,
1:05:38equipment that what looks like a scaled
1:05:40up version of what people are using on
1:05:42the bench. Maybe we we can do something
1:05:44different. And and this is this is where
1:05:46where we we came from. We've been
1:05:49working in continuous flow manufacturing
1:05:52technologies for many years
1:05:55for advanced materials as well as other
1:05:58types of
1:05:59drug. And the challenge here, of course,
1:06:02is compared to to other systems is that
1:06:07you know, synthesizing RNA, making RNA
1:06:10does take quite a long time in
1:06:12comparison. Just to give you an idea,
1:06:15you know,
1:06:16using a continuous flow system to make,
1:06:19I don't know, some inorganic
1:06:20nanoparticles, you might require
1:06:23literally a couple of seconds,
1:06:24literally, and and we've done that. And
1:06:27whereas here, you're looking at a few
1:06:29hours. So so really somebody had to
1:06:31think completely outside of the box to
1:06:33make this work. And so we did. Whole
1:06:36thing started quite serendipitously, but
1:06:38then then we we made it work.
1:06:41And what we realized was once you you
1:06:44you have you bring this new type of
1:06:47thinking, you bring these new
1:06:48technologies in, then that opens up a
1:06:51lot of opportunity actually to innovate
1:06:54and innovate how how these drugs can be
1:06:56delivered to the patients. So instead of
1:06:58having these large centralized
1:06:59facilities, now we were able to
1:07:01actually, you know, miniaturize
1:07:04manufacturing, make it at appropriate
1:07:06scale, right, make it at at batches that
1:07:09are completely configurable depending on
1:07:12the what the the particular target is,
1:07:15the particular therapeutic target is.
1:07:18And and it would not have been possible
1:07:20with with any conventional technologies,
1:07:21we felt. So so we we gave it a a really
1:07:25really good crack there the the last,
1:07:27you know, 5 years or so and we ended up
1:07:31with with platforms that that are
1:07:32actually, you know, now productized and
1:07:34they're out there.
1:07:36So, so, you know, with this new thinking
1:07:39we're looking at
1:07:41really, you know, a future where
1:07:44personalized treatments are going to be
1:07:47very very possible as close to the
1:07:50bedside as as as possible, okay? So,
1:07:54what was needed was the infrastructure.
1:07:57We have developed the infrastructure and
1:07:59of course with the the ARPA-H contract
1:08:02now, which we're very excited about
1:08:04this.
1:08:05This, you know, not only broaden
1:08:08the scope
1:08:09>> [laughter]
1:08:11>> We're sorry about it, too.
1:08:14Ben's got trigger finger on the on the
1:08:16soundboard, guys. We apologize.
1:08:17[clears throat and laughter]
1:08:20>> I love it.
1:08:22>> [laughter]
1:08:23>> So, so, yeah, not not only, you know,
1:08:26bringing um
1:08:28providing the the ability to bring these
1:08:29technologies close to to the bedside, uh
1:08:32but also um um help us
1:08:35uh you know, broaden the scope of of the
1:08:37technology itself. So, we're looking now
1:08:41at uh complete automation um um
1:08:44including uh fill-finish, which is
1:08:47remains a massive massive bottleneck.
1:08:49So, imagine, we're talking about
1:08:50patients of one, right? How do you
1:08:52actually uh take uh the bulk drug
1:08:55product which you have made in our
1:08:58technology, using our technology or
1:08:59somebody else's, it doesn't really
1:09:00matter. But again, and bring this uh
1:09:02close to the bedside. So, this is a a
1:09:04very big technical challenge still.
1:09:07Yeah, and then there's other things like
1:09:08advanced measurements, uh this
1:09:10automation as I mentioned, and data
1:09:13traceability, and and you know, the
1:09:16analytics, can we actually move towards
1:09:19a real-time release scenario, which is a
1:09:21little bit down the road, but but this
1:09:24this contract gives us uh these
1:09:26opportunities to to explore these
1:09:28things. So, we're looking at medicine uh
1:09:33becoming more and more programmable.
1:09:36Um and by programmable, what we mean is,
1:09:41you know, um deliver against uh needs of
1:09:44individual patients.
1:09:46>> Mhm.
1:09:47>> But, the manufacturing infrastructure
1:09:49should must be programmable as well,
1:09:52okay? So, when batch manufacturing, the
1:09:56very conventional manufacturing methods
1:09:58were designed around the factory,
1:10:01what we're delivering right now, uh
1:10:03especially through the RPI gift program,
1:10:06we're delivering, you know, these new
1:10:08ideas, continuous flow manufacturing in
1:10:10particular, and all the other sort of um
1:10:13uh um um
1:10:15uh related components, um to design
1:10:19manufacturing around the medicine
1:10:21itself, and
1:10:22in particular, the patient. And that
1:10:25makes us extremely excited.
1:10:28>> Excellent. Anything to add to that, uh
1:10:30Peter?
1:10:31>> [clears throat]
1:10:31>> Yeah, I think uh I mean uh Errik
1:10:34definitely explained it quite nicely,
1:10:35but I guess I'll take a bit of a view
1:10:37from the analytical perspective, uh
1:10:39since that's um you know, Base Life
1:10:41Sciences really their strong expertise.
1:10:44Um
1:10:45I think overall, um you know,
1:10:47um the manufacturing that currently is
1:10:49set up is large scale, uh so you need
1:10:51large volumes. Uh there's typically, you
1:10:53know, a ton of different samples that
1:10:55you need to take along the process, um
1:10:58as well as the final product, uh that
1:10:59you need test need to test with
1:11:01you know, a large variety of different
1:11:03methods that are all have their
1:11:04independent machines and equipment. Um
1:11:07so again, uh not only the manufacturing
1:11:09facility itself is large, but then also
1:11:11the analytical facility that is attached
1:11:14to it is also large. Um
1:11:15>> Right.
1:11:16>> Um so, that is really uh a challenge uh
1:11:19to keep these things at a cost that is,
1:11:21you know,
1:11:22uh makes makes these drugs available for
1:11:24for the people that need it. Um
1:11:26So, with this program here, we're really
1:11:28hoping to
1:11:30incorporate that as much as possible.
1:11:32Um, it will be a challenge, um, but um,
1:11:35hopefully it will lead to a
1:11:37more standardized process, um, you know,
1:11:40where the FDA can say, "Hey, yeah, sure.
1:11:42Our regulatory guidance is on the
1:11:44process itself and not on all the
1:11:46individual components."
1:11:48Um, so that's that will be, um, you
1:11:50know, a major breakthrough,
1:11:52um, and um,
1:11:53you know, we're as PACE will bring sort
1:11:55of the the guidance with our current now
1:11:59standardized, uh, um, methods that we
1:12:01use across many manufacturing runs
1:12:04and then see if that translates directly
1:12:06to this to this machine.
1:12:08So,
1:12:10yeah, that's really where we where we
1:12:11see the challenge, I guess, in the
1:12:12current setup and where we, uh, are
1:12:16pretty confident, I think, with the team
1:12:18that we have, um, because we have a
1:12:20pretty excellent team, uh, that we can
1:12:22address these challenges and that we
1:12:23can, you know, satisfy regulatory needs
1:12:26as well as, uh, you know, the time and
1:12:28the cost savings that that we're going
1:12:30to see.
1:12:31>> Excellent. Perfect. Well, ladies and
1:12:33gentlemen, if you have any questions,
1:12:34feel free to put them in the chat. We
1:12:35can get to them. If you have any
1:12:36questions for Harris or for Peter. Um,
1:12:38but yeah, just I guess to to set the the
1:12:40stage for this again, right? Which is I
1:12:42think just so cool about what you guys
1:12:44are working on and then also just like
1:12:46the whole ARPA-H program as well.
1:12:48The overall [snorts] goal here is
1:12:50basically we have all of these
1:12:52personalized diseases where there's like
1:12:54a specific difference for each
1:12:55individual person. We cannot simply make
1:12:58a one-size shoe that's going to fit
1:13:00everybody. We've got to make individual
1:13:02therapies for each person. How in the
1:13:04world do we make millions of
1:13:06individualized doses eventually? And so,
1:13:10this is kind of where I'm so excited
1:13:12about what you guys are are building.
1:13:14And so, I would I would love to kind of
1:13:15get your, your your story about how this
1:13:18kind of came to be. Like, when how did
1:13:20you first get into this, Harris? Um,
1:13:23what was kind of like the the key unlock
1:13:25with regards to realizing, "Oh, shoot.
1:13:28We've got a a
1:13:29stack here that can make personalized
1:13:32doses at scale.
1:13:36>> Yeah, okay. So, so again, I mean this is
1:13:38uh this is a very very important
1:13:39question. And and this is something we
1:13:41have spent
1:13:43a very very long time actually um
1:13:45thinking about and and working on in
1:13:47fact. So, we we came from a completely
1:13:51different angle when we started doing
1:13:54that work back in 2017 and later on with
1:13:58uh you know, our partners uh Welcome
1:14:00Leap
1:14:01who actually funded this this initial
1:14:03piece
1:14:04together with the with the UK
1:14:06government. And we we came from from an
1:14:09angle of vaccine manufacturing. So,
1:14:12we're we're talking about mass market,
1:14:13right?
1:14:14>> Yeah.
1:14:15>> Right.
1:14:15>> Uh but then
1:14:17one of the the key requirements in these
1:14:20programs was how this technology or how
1:14:24a technology could be um designed such
1:14:27that it can be deployed as close to the
1:14:31patient as possible.
1:14:33And especially thinking low and
1:14:35middle-income countries.
1:14:36So, so we we we we started engineering
1:14:40from from this requirement. But then
1:14:41then we realized, okay,
1:14:44the we had the pandemic. We we had to
1:14:46have a very fast response. Fast response
1:14:49means what? Okay, you have to make at
1:14:50scale obviously when you get to that
1:14:52stage. But in order to get to that
1:14:54stage, you you cannot immediately start
1:14:56making, you know,
1:14:58uh liters of this of this material,
1:15:00right? You you have to start small.
1:15:02So, so the the approach we have taken
1:15:05was or the question we tried to answer
1:15:07was, can we have exactly the same
1:15:09platform using this manufacturing
1:15:12technology
1:15:14uh such that we can do small-scale work
1:15:18for discovery
1:15:20or process development and then use the
1:15:23same platform to scale it.
1:15:26That was the engineering challenge uh we
1:15:28we we set ourselves initially. And the
1:15:32thing was, okay, if our continuous flow
1:15:36uh, you know, methodologies
1:15:39could actually do what they said on the
1:15:41tin,
1:15:42then that could be well possible.
1:15:46So, we spend a lot of engineering time
1:15:50to actually address the problem. And the
1:15:53the way we have addressed it was keep
1:15:55the process exactly the same, the
1:15:57methods exactly the same,
1:15:59but then develop
1:16:02three sized consumable cartridges. Our
1:16:05technology is based, you know, is is
1:16:07working on on consumable cartridges, and
1:16:09that ticks a major regulatory box,
1:16:12especially, uh, you know, for for
1:16:14clinical supply and and production.
1:16:16And we thought, okay, we're going to
1:16:17have uh, literally the small, the
1:16:19medium, and the large. The large being
1:16:22for large-scale, you know,
1:16:24country-level,
1:16:26whatever, you know, scale. But then you
1:16:29have the the small for discovery and the
1:16:31medium for process development. So, with
1:16:33this, um,
1:16:34in mind, we actually built it, we tested
1:16:37it, it worked perfectly well. We
1:16:38thought, okay, so I mean, the next
1:16:41pandemic is is is not going to happen
1:16:43anytime soon, right?
1:16:47>> Don't [laughter] Don't jinx it. Please
1:16:48don't jinx it.
1:16:51What have you done?
1:16:52>> Completely Completely So, look, I mean,
1:16:54we we we we from a very very very very
1:16:55very
1:16:56sort of
1:16:56>> [laughter]
1:16:57>> positive, you know, angle, very positive
1:16:59position, okay? It's not going to
1:17:00happen.
1:17:01Uh, but if it does, I mean, we have the
1:17:03the tools uh, now to to to to to
1:17:05>> He's thinking, "Ching ching. Here we
1:17:07go."
1:17:07>> [laughter]
1:17:08>> It's not very frequent. So, so yeah, so
1:17:11so we had to we had to actually think of
1:17:14of other, you know, application areas.
1:17:15And since we could do it, then this is
1:17:18the obvious one. So, the approach we're
1:17:20taking, uh, and and again, this is going
1:17:23to happen with uh, the ARPA-H
1:17:25give award.
1:17:27Um,
1:17:28we are going to enable a distributed
1:17:31manufacturing. So, as close to the
1:17:32bedside as possible. And that's where
1:17:35also Pace come in. In this
1:17:39contract, Pace are our deployment
1:17:41partner. They have sites throughout the
1:17:43the United States. This is extremely
1:17:45important.
1:17:47And a very, you know, very real world,
1:17:49real life, you know,
1:17:52test of of of this idea.
1:17:55And so that's that's one, you know,
1:17:57distributed manufacturing as close to
1:17:58the bedside as possible. And number two,
1:18:01have a machine
1:18:04that does deliver the entire technology
1:18:06stack, start with literally dialing in
1:18:10the gene
1:18:11in a computer.
1:18:14Out comes the the the drug product ready
1:18:18to to release and administer and stuff
1:18:20like that. But then, you know, the the
1:18:23other engineering challenge which we now
1:18:26have to to to to address,
1:18:28and we do have very solid,
1:18:31you know, route to actually addressing
1:18:33this problem, how we can make in
1:18:37parallel,
1:18:38you know,
1:18:39um
1:18:41individualized drugs
1:18:43for as many patients in a day as
1:18:46possible. We have set ourselves a
1:18:47target, single machine to be or single
1:18:50instance of of this the technology we're
1:18:52developing right now,
1:18:53to to be able to to deliver for 100
1:18:57patients per day. Okay? So, distributed
1:19:00manufacturing, 100 patients per day, one
1:19:03instance of the technology. I think we
1:19:06we are on very very,
1:19:08you know, on the right tracks to
1:19:10actually address this scalability
1:19:12problem. Now, the the other challenge,
1:19:15of course, and and I'm sure our
1:19:18friends from the regulatory authorities
1:19:21might be listening to this, how do we
1:19:23make this regulatory compliant? That is
1:19:25extremely important. We already have
1:19:27work streams in there. And of course,
1:19:29that's also where Pace
1:19:31come in
1:19:33to actually make sure that that, you you
1:19:36whatever comes out of of system is 100%
1:19:40regulatory compliant
1:19:42and there's no compromises there
1:19:44whatsoever.
1:19:45>> Yeah. Okay, so maybe this is where you
1:19:47can jump in as well. Uh uh
1:19:49uh
1:19:50Peter, there's a question from from Ali
1:19:52Oop. Um how do you gauge safety and
1:19:54efficacy for these personalized
1:19:55therapies, right? You're you're making
1:19:57these things at scale. What's kind of
1:19:59the the framework uh that you build in
1:20:01behind from an analytical perspective
1:20:03and release testing for these things?
1:20:05>> Yeah, I think maybe to um follow up a
1:20:08little bit on on um
1:20:10Harris's uh comments just now, I think
1:20:12it's a bit more of a philosophical uh
1:20:14again different approach uh to all of
1:20:17this what we're developing here.
1:20:19Um it's really developing a
1:20:20manufacturing
1:20:22platform, QC platform that is always the
1:20:24same um that we can use for all these
1:20:27different um you know,
1:20:29personalized medicines that are under
1:20:30development right now.
1:20:32So, I guess the
1:20:33traditional approach is like hey, we
1:20:35have a target that we want to
1:20:37um
1:20:38you know, make medicine against using
1:20:39AAV vector or or an LNP or mRNA.
1:20:43Um so, first [clears throat] we develop
1:20:45our our asset, make sure that it works
1:20:47in vivo models and whatnot, you know,
1:20:50make sure that's safe, go into the
1:20:51clinic after we produce a batch that is,
1:20:54you know, a a platform then, a
1:20:56manufacturing platform
1:20:58that you're designing around your asset.
1:21:00Um so, right now we're designing a
1:21:01platform
1:21:03uh that any asset can go into as long as
1:21:05a mRNA or LNP based.
1:21:07Um so, I think from that angle and this
1:21:09goes back to the question that was just
1:21:11asked.
1:21:12Um you know, we are
1:21:14um ideally at the end of this entire
1:21:16development program in a position where
1:21:18we can say anything that goes in and out
1:21:20of this machine is fully quality
1:21:22controlled and we know that what comes
1:21:23out of it is always good. Uh the potency
1:21:26of it, you know, obviously needs to be
1:21:28tested. This is really where the drug
1:21:30developers come in and we have excellent
1:21:31drug developers in the program as well.
1:21:33So, the uh the end goal would be to test
1:21:36at least three of those into real drug
1:21:40products that can be tested in in
1:21:42clinical trial.
1:21:43Um
1:21:44So, yeah. So, it's a bit of a
1:21:45philosophical difference in a different
1:21:47approach, I guess.
1:21:49Yeah, where you really the platform is
1:21:53>> [clears throat]
1:21:53>> uh is sort of the the you know the the
1:21:56the the glue that holds it all together.
1:21:58Uh and all these different modalities or
1:22:00all these different assets with you know
1:22:03it's really just a a genetic code
1:22:05uh that for a personalized medicine that
1:22:07will be different from patient X to
1:22:09patient Y uh even though it's sort of
1:22:11the same disease, but they just have
1:22:13different mutations. So, we can make
1:22:15that in this machine now, and that's why
1:22:16the N of 1 is so important. Uh so, we
1:22:19make a small batch just for patient X.
1:22:22They'll be treated. You know, in the
1:22:23afternoon we make the the drug for
1:22:25patient Y, and they will be treated uh
1:22:27basically on the same day, but it's
1:22:29still a different batch. Um and the
1:22:31quality control and the manufacturing
1:22:32processes are exactly the same.
1:22:34Uh and this will hopefully then lead to
1:22:37sufficient trust um and confidence by
1:22:41the public, by the FDA,
1:22:43uh to regulatory um agencies to assure
1:22:46that this is uh you know that we can
1:22:48that we can move forward with this.
1:22:50>> So, I have a question about just like
1:22:52just in general. I think I'm thinking of
1:22:54like the personalized mRNA vaccines from
1:22:57like Moderna, right? If we can if we can
1:22:59design these things in a matter of days,
1:23:02but it still takes like weeks or months
1:23:05to actually get the thing to the
1:23:06patient, how much of that is the
1:23:08manufacturing um versus like everything
1:23:11else, the the design, uh the
1:23:14you know, everything else. Um I guess
1:23:17the the bottleneck that you're solving
1:23:19is it what is it the biggest bottleneck?
1:23:22Or I guess the tightest bottleneck, I
1:23:23guess.
1:23:25>> It's a large bottleneck, for sure,
1:23:27because the manufacturing currently
1:23:28takes, you know, weeks, and then you
1:23:30have your release, which takes other
1:23:32more weeks. Um so, and then I guess
1:23:36before you even get into a manufacturing
1:23:38spot, you have to reserve it months in
1:23:39in advance.
1:23:41Um and then you make way too much for
1:23:42these personalized medicines, so it's
1:23:44going to be very expensive.
1:23:46Um so yeah, [clears throat] it is um
1:23:48it's a large bottleneck. It's not the
1:23:49only one, of course. I mean, I guess
1:23:51personalized medicines are
1:23:53essentially in their infancy, really, uh
1:23:55still, gene therapies. Um so we still
1:23:57need to validate a lot of these
1:23:59uh targets. So, you know, a personalized
1:24:02cancer
1:24:03um vaccine
1:24:05is great as long as we have the right
1:24:07epitopes that we're targeting. So, we
1:24:09need to learn a lot more about that.
1:24:11Um but this new machine will also um
1:24:14speed that up because now we can make in
1:24:17a day or in a week, uh you know,
1:24:19multiple batches for several different
1:24:21patients that all have their unique um
1:24:23genetic mutations.
1:24:25So, we can go into the clinic way
1:24:26quicker. We can validate it quicker.
1:24:29Uh and then we know, okay, this is
1:24:30indeed the right target that we want to
1:24:32address.
1:24:33Um and we can, you know, move forward
1:24:36other patients uh that we can then
1:24:37produce very quickly uh based on their
1:24:40genetic mutations.
1:24:41Yeah, so you think that?
1:24:43>> Yeah, I think I think Peter, you you
1:24:45have explained it really, really well.
1:24:47So, um the there are, of course,
1:24:50bottlenecks throughout, for sure.
1:24:53And uh logistics is also a bottleneck,
1:24:56you know, from from biopsy to actually,
1:25:00uh you know, uh creating the uh the the
1:25:02sequence and testing it and then making
1:25:04it and so on and then taking it back to
1:25:06the patient
1:25:08uh is I mean, this this entire entire
1:25:10chain is is a bottleneck.
1:25:12And and typically, at least in in in our
1:25:15experience, the chemistry and
1:25:16manufacturing controls, the CMC, has
1:25:18always been the major bottleneck, and
1:25:20that's where a lot of the the sort of
1:25:23the cost and the value lies, also,
1:25:26right? So, so
1:25:30well, what we have achieved so far, just
1:25:32just to give you to give you because we
1:25:34we we're not starting from a blank sheet
1:25:36of paper here, okay? So, so what what
1:25:38what we're claiming is is based on on on
1:25:40some real world experience we have.
1:25:43Um the installations we we're doing
1:25:45right now,
1:25:46And somebody or the the recipients of
1:25:49the technology
1:25:52are in a position to make 10 completely
1:25:54different products per week and release
1:25:58them. Okay?
1:25:59And and that is
1:26:01at least phase one
1:26:05you know level of
1:26:07quantity and quality of of material.
1:26:10So
1:26:11to from where we are today and of course
1:26:14that they can progress that they get
1:26:15into GMP and you know sorry scale it up
1:26:18and and and do whatever. So from where
1:26:20we are today
1:26:22to the target we have set in in this new
1:26:26contract this this new new project
1:26:29is
1:26:30is challenging. It is extremely exciting
1:26:34but it is not completely wacky. It is it
1:26:37is very very real and realizable within
1:26:41the the time frames. Now
1:26:43um
1:26:44regulator compliance is very important
1:26:46and as our our colleague on on the chat
1:26:49mentioned you know testing for for
1:26:52efficacy and all that stuff is is
1:26:55important. Okay? So
1:26:58there's two levels to that.
1:27:00Of course from from a CMC from a
1:27:02manufacturing perspective what what can
1:27:04you do? You know you have the
1:27:05manufacturing stuff you know you can
1:27:07make you you you you know you can prove
1:27:09you can make at at a certain level of
1:27:12quality hopefully high quality. Then you
1:27:15know you you know you can analyze and
1:27:18this is a key component in in this
1:27:20particular program but also it it is up
1:27:23to the the designers as well
1:27:25to actually you know design
1:27:28for the particular targets they are
1:27:31they're aiming at.
1:27:33Having said that still having platforms
1:27:36that allow them to iterate fast fail
1:27:40fast
1:27:42also shortens the time to the patient.
1:27:45So that's another angle. Okay? And
1:27:49I mean
1:27:50it is and it isn't in the personalized
1:27:52space. It is a much broader
1:27:55sort of
1:27:56use case as it were, much broader
1:27:58advantage
1:28:00which somebody can address with with
1:28:01these new new technologies for sure.
1:28:05>> Excellent.
1:28:05>> [clears throat and cough]
1:28:06>> Okay, so I know we've only got a couple
1:28:07minutes left here. I want to be super
1:28:09respectful of both of your
1:28:11both of your folks' time.
1:28:12So So two questions I guess just quickly
1:28:14here as as we wrap up. So we'll need
1:28:16some rapid fire responses.
1:28:18ARPA-H, I love ARPA-H. I love the whole
1:28:21idea of pragmatic moonshots and the way
1:28:24they they structure their their their
1:28:25their contracts. What was what was like
1:28:28the the elevator pitch to them that just
1:28:30kind of made them like sit forward and
1:28:31be like, "Oh shoot, yeah, we need to we
1:28:33need to give these guys a contract."
1:28:36>> Um Peter, did you want to go first or
1:28:39>> Uh why why don't you go?
1:28:41>> All right, okay. So
1:28:43it was
1:28:44the team absolutely that the consortium
1:28:47put together. I'm I'm very proud about
1:28:49about this. So for me this is this was
1:28:51number one. And everyone brings very
1:28:54crucial components into this. And B, we
1:28:57don't start from scratch. So there is
1:28:59there was a very clear clearly specified
1:29:02you know requirement and we're able to
1:29:04address you know all of the requirements
1:29:08ARPA-H has has set in in the challenge.
1:29:11And the other thing of course, you know,
1:29:12it is absolutely amazing that the team
1:29:15the gift team are you know, they they
1:29:18are really really world-class experts.
1:29:20It is absolutely
1:29:21I'm very impressed. Absolutely amazing
1:29:23to work with them. Really really joy to
1:29:25work with them.
1:29:27>> Excellent.
1:29:28>> So from [clears throat] base
1:29:29perspective, I think what was really
1:29:32enticing for ARPA-H is that we are sort
1:29:34of rooting this program in reality
1:29:38with you know extensive expertise
1:29:40[clears throat] in in you know current
1:29:42release methods. So we can really show
1:29:45that what comes out in the machine that
1:29:48we're going to develop uh, holds true to
1:29:50to what is the current benchmark. So,
1:29:53that's very exciting, um, and uh, and
1:29:55and enticing for ARPA-H. The other
1:29:58thing, I mean, I'll just
1:29:59uh, you know, echo what uh, what Harris
1:30:02said here is that when we went to
1:30:04Pittsburgh for the kick kick-off
1:30:05meeting, uh, and the entire team saw
1:30:07that there was actually a device that we
1:30:10are starting with,
1:30:11>> [laughter]
1:30:11>> uh, which was was just amazing, and it
1:30:13was was very cool to see. Uh, and we're
1:30:15really super excited to just bring this
1:30:17forward and develop a machine that
1:30:20for real, you know, in the near future,
1:30:22we will be able to deliver genetic
1:30:24medicines, individualized genetic
1:30:26medicines to everyone that needs it, so.
1:30:28>> I love how the big differentiator is
1:30:30ours is living in reality.
1:30:32>> Right.
1:30:32>> [laughter]
1:30:33>> It is.
1:30:34>> shows how how like crazy things have
1:30:37gotten sometimes. [laughter]
1:30:38>> Yeah.
1:30:39>> But, no, it's that's really cool. And I
1:30:40love how Ben's favorite thing is the
1:30:43US-specific ARPA-H award,
1:30:46being a Canadian.
1:30:47>> [laughter]
1:30:48>> I can sense your jealousy, Ben.
1:30:52>> Carney, for crying out loud, do
1:30:53something similar.
1:30:54>> [laughter]
1:30:55>> Um,
1:30:55excellent. Well,
1:30:57well, ladies and gentlemen, it's been a
1:30:59pleasure to have Centellian and and Pace
1:31:01on today. Thank you so much, folks.
1:31:02Thank you both to Harris and to Peter
1:31:04for coming on.
1:31:05Very excited to see where you guys take
1:31:07this, and thanks so much. We look
1:31:09forward to following your progress.
1:31:10>> Thank you very much. It's been really
1:31:12exciting.
1:31:12>> Cheers.
1:31:12>> All right, thanks so much for coming.
1:31:14>> Thanks, you guys.
1:31:15>> All right, guys.
1:31:16>> Right on. All right. So,
1:31:20without any further ado, we are going to
1:31:21be
1:31:21>> having a bunch of RNA therapeutic
1:31:23companies this week.
1:31:24>> Apparently, yes.
1:31:26>> Yeah. Really good.
1:31:27>> Anyway, that being said, um, we're going
1:31:29to be moving on to our next our next
1:31:31guest, uh,
1:31:32Ryan uh, Coward. He's here in the in the
1:31:35in the in the studio as well. Let's
1:31:36bring him on without any further ado.
1:31:38Just a second here. All right, Ryan.
1:31:41>> sponsor.
1:31:42>> Yes, excellent. Ryan, welcome to the
1:31:45show. Good to see you again.
1:31:47>> How's things? How's it going?
1:31:48>> Doing well. Good to see you again.
1:31:50>> Good, good. Excellent. Show's looking
1:31:52good.
1:31:53>> Oh, well, thank you. Yeah, it's been
1:31:54it's been a wild last couple weeks
1:31:56inside of biotech.
1:31:58I I actually one of the things we didn't
1:32:00get to on the on the last show and
1:32:01actually I might be interested to get
1:32:02you to to chime in on it a little bit
1:32:05later on as well, Ryan, would be like
1:32:06the the the new draft guidance from the
1:32:08FDA with regards to personalized
1:32:10therapies. I don't I'm I might be
1:32:12springing it on you and if you want to
1:32:14talk about something else, that's fine.
1:32:15>> Well, I have to say you are springing it
1:32:17on me. I could I could I could guess.
1:32:19You could tell me the outline, but yeah.
1:32:21>> [laughter]
1:32:21>> All right. We can get to that a little
1:32:23bit later. Excellent. Okay. Well,
1:32:26Ryan, I was I was going I was going
1:32:27through your through through your
1:32:29background and and and you spent a
1:32:31obviously a lot of time at at Oxigene.
1:32:34I'm I'm curious like as you as you kind
1:32:36of look into the the world going forward
1:32:38from from building out like a biotech
1:32:41company,
1:32:43what would what would kind of be like
1:32:44your recommendations given like the the
1:32:46the
1:32:47the
1:32:48the the mistakes in the and the and the
1:32:50and the drawbacks that can happen in
1:32:52building out a biotech company. Like
1:32:53what would be some of the things that
1:32:54you're seeing now that you think will be
1:32:56really effective for the next generation
1:32:58of founders? Ideally like either tools
1:33:00that they can use or mistakes that they
1:33:01can avoid.
1:33:03>> I think
1:33:04I know there's quite a lot of mantras
1:33:06I'd say and and sort of, you know, I
1:33:08suppose in startups more generally in
1:33:09tech it's not just biotech, right?
1:33:11There's a lot of these mantras around
1:33:12like
1:33:13you know, raise everything you possibly
1:33:14can right now and you know, because you
1:33:17might be only be able to raise it now
1:33:18and it's like I've always I don't
1:33:20necessarily fundamentally disagree with
1:33:22that as a as a mantra, but I think it's
1:33:24it's a slippery slope and slightly
1:33:25dangerous in the sense that, you know,
1:33:28you sort of you can end up in that sort
1:33:29of Theranos territory, right? Where
1:33:30you've got so much money the product has
1:33:32to deliver. Like my view has always been
1:33:34like raise money that is proportional to
1:33:35your technology at a share price you're
1:33:37comfortable with. Like this,
1:33:39you know, it's very easy to just say
1:33:40raise everything that you possibly can.
1:33:42Um
1:33:43So that's kind of one of those things I
1:33:44think that I've I've I've come across a
1:33:46lot as like a generalization that I'm
1:33:48I'm not I don't entirely necessarily
1:33:50agree with and and it's a really baby
1:33:52one of those other things is like you
1:33:54have to just be working all the time
1:33:56endlessly, otherwise you're not doing it
1:33:57right. And it's like I'd say most of my
1:33:59strategic thinking has come in the bath
1:34:02reading a book.
1:34:03>> [laughter]
1:34:04>> You know, I'm like, "Oh, oh wow, I
1:34:05realize I now need to fundamentally
1:34:07change the direction of the company."
1:34:08And it's
1:34:08>> [laughter]
1:34:09>> It's not It's not because I was
1:34:10tirelessly working at 3:00 a.m. in the
1:34:12morning, you know, doing stuff because
1:34:14it needed to be done. It was actually
1:34:15just thinking strategically about the
1:34:17business. So, you know, I think um
1:34:19>> Normally you're fixing Normally you're
1:34:20fixing problems that you made at 3:00
1:34:22a.m. while working tirelessly.
1:34:24>> Yeah.
1:34:24>> [laughter]
1:34:24>> That's true. Yeah.
1:34:25And it's not to say I still didn't have
1:34:27to continue to work at 3:00 a.m. on
1:34:28many, many days, but like, you know, I
1:34:30think there's a There's kind of that
1:34:32mantra that like you have to just grind
1:34:33away endlessly and that's the key to
1:34:35success. And I would say it's like
1:34:36strategy and strategic thinking is
1:34:38probably more valuable than just doing
1:34:39stuff. If that makes sense. So, yeah,
1:34:42maybe that Maybe that could help in
1:34:43terms of advice.
1:34:44>> Well,
1:34:45>> [clears throat]
1:34:45>> I don't want to jump in and like derail
1:34:47conversation Ben, but I I do want to ask
1:34:51about [clears throat]
1:34:52the first story we covered today was
1:34:53Mythril. Uh it's about a company that
1:34:56was bought by Anthropic. Ben's been on
1:34:58this soapbox for over a year now that
1:35:02biotech is being taken over by the tech
1:35:04bros.
1:35:05Yeah. And we're going to see more of
1:35:07these one-man
1:35:10biotech companies that just use Claude
1:35:12code and their fingertips and a keyboard
1:35:15to to design some therapeutic. Coming
1:35:17from a guy
1:35:19who's been in the space for a long time.
1:35:21Now you're in software. What do you
1:35:22think of Ben's opinion?
1:35:24>> [laughter]
1:35:25>> Uh
1:35:26Do I want to disagree with that? Um no,
1:35:28I think I think like
1:35:30>> Pick a side, Ryan. Come on.
1:35:32>> Yes. [laughter]
1:35:33That's That's no No halfway house there.
1:35:35Um
1:35:36yeah, I I would say that like what has
1:35:39been fundamentally transformational is
1:35:40the ability to make a piece of software
1:35:42for a particular application. So, like
1:35:44if you say like I need to do this
1:35:45particular thing, you can spin that up
1:35:47relatively easily. Um so, that you can,
1:35:50you know, you have your software to
1:35:51solve that particular problem. Um
1:35:53that's fine and it works well in that
1:35:56context. But if you're like, "Actually,
1:35:57this needs to be connected to four or
1:35:59five other pieces of software and you're
1:36:00trying to spin those up and make them
1:36:01work synchronously together,
1:36:03things start to break down relatively
1:36:05quickly and I think you do still need
1:36:06that here sort of human interaction and
1:36:07people with software skills to be able
1:36:09to say actually this is the way you lay
1:36:10down a complex enterprise level piece of
1:36:12software. But if you're a scientist and
1:36:14you just you need to solve a particular
1:36:15problem then yeah, it's certainly within
1:36:17your wheelhouse to to do that now. But I
1:36:19wouldn't necessarily say that that is a
1:36:20software product.
1:36:22That is solving a particular scientific
1:36:23problem with a piece of software you
1:36:25just created for that need.
1:36:27>> Yeah, I hadn't thought of like
1:36:28>> I think I just found the middle ground
1:36:29there, right? I found the middle ground.
1:36:31>> Well, I I and I hadn't I hadn't really
1:36:32thought of like the AI we we think of
1:36:35like these AI coding agents as being
1:36:36really good at making software. Um but
1:36:38in a certain sense though, I think that
1:36:40biology is kind of like software. It's
1:36:42sort of Yeah.
1:36:44>> Yeah, I mean it's really it's really
1:36:47messy 3 billion year old largely
1:36:50unoptimized code.
1:36:53>> [laughter]
1:36:54>> I mean it's it's a software guy's
1:36:55nightmare.
1:36:58>> [laughter]
1:36:58>> Which makes it good for AI.
1:37:01>> Yeah, no exactly and I think it is
1:37:02really good for that. Um but you know,
1:37:04just finding patterns and things like
1:37:05that. But you know, the original AI
1:37:06algorithms were designed to work on
1:37:08like, you know, blocks of text. But DNA
1:37:10doesn't really have blocks, right? It's
1:37:11just a continuous string that's billions
1:37:13of bases long. And so it you
1:37:15contextually it's not quite um designed
1:37:17in the way that it is, you know, large
1:37:19language models are designed to work on
1:37:20language um which is broken up
1:37:23discretely into meaningful contexts, but
1:37:25DNA doesn't really have that. So um you
1:37:27know, we can identify promoters, we can
1:37:28identify particular sequences, but um
1:37:31it's not it's not designed to really
1:37:33read long continuous strings like that.
1:37:35So, you know, but it's still great.
1:37:38>> Fair enough. Um so yeah, kind of I guess
1:37:41like to some extent um walk me through
1:37:44then, right? As you start to build out
1:37:46an organization where I I I I agree. I
1:37:49think for something like Mithril, um
1:37:51there are
1:37:54circumstances at the early stage where
1:37:56you can keep the company as one person.
1:37:58Maybe you can maybe and maybe you can
1:38:00keep it a fairly lean team. Um, but as
1:38:03you as you start to build out something
1:38:05obviously that's far more complex,
1:38:06you're going to just need more people to
1:38:08to manage everything.
1:38:09>> Yeah.
1:38:09>> Um, and so kind of just like like look
1:38:12looking back over your career, uh, what
1:38:14would you say are some of the things
1:38:15that work fine when you're a five-person
1:38:18team or smaller but start to break as
1:38:19you scale towards 50?
1:38:22>> Uh,
1:38:23everything breaks as you go from
1:38:27I mean, there is not there is nothing
1:38:28that works well in a five-person company
1:38:31that will scale to a 50-person company.
1:38:33Like the reality of the circumstance is
1:38:35very different, you know, you're in a
1:38:36room, you can speak to these people
1:38:37every day. The project management There
1:38:39is no project management. Like the
1:38:40project management is the desk with the
1:38:41guy next to you and the and the
1:38:43scientist next to you. It's you have
1:38:45that fluid dynamic chat. Um, and
1:38:47everything everybody knows everything.
1:38:49Everybody knows everything that's going
1:38:50on and whether they need to or not, they
1:38:52know it. Whereas as soon as you get to
1:38:5450 people, you've got, you know,
1:38:56different tiers within the organization.
1:38:58You're no longer as a as a leader
1:39:00operational, you're strategic. So, you
1:39:01know, if you're still sweeping the
1:39:02floors when the company is at 50, you're
1:39:04doing it wrong. So, you've got to go
1:39:05through that transition and that
1:39:06journey. Um, which you know, going back
1:39:09to the what you were asking me earlier
1:39:10about like, you know, the the sort of
1:39:11stereotypical things that people say
1:39:13like the you know, some people say, "Oh,
1:39:15the see the the scientist CEO should be
1:39:17replaced at some point by a professional
1:39:19CEO." And I'm like, it's
1:39:21not necessarily, right? It's a
1:39:22transition and that's that person either
1:39:24has the ability to make that transition
1:39:25from the five-person company where
1:39:27they're very operational to the
1:39:2850-person company where you're thinking
1:39:30largely about strategy because if you
1:39:32are doing the operational stuff, you're
1:39:33probably not running the company
1:39:34effectively because, you know, you you
1:39:36shouldn't be doing that. So, but
1:39:38everything breaks when you go from five
1:39:40to 50. Like, you know, you you need
1:39:41people's expectations of what the
1:39:43company is in terms of infrastructure,
1:39:45in terms of the support they get and you
1:39:47know, you need a HR department. You
1:39:48don't need that when you're five people.
1:39:49Like, it's like all your file structures
1:39:52and everything break. It's, you know,
1:39:53conflicting copies left, right, and
1:39:55center. People can't understand what's
1:39:56going on in different parts of the
1:39:57building. Like, it's yeah. It No nothing
1:40:00at five survives to 50 is my point.
1:40:02>> I've even seen nothing that you have at
1:40:04one survives with two people.
1:40:06>> [laughter]
1:40:07>> That's what I'm going through right now.
1:40:09>> I don't know if this is insight to your
1:40:11RELATIONSHIP OR
1:40:12>> [laughter]
1:40:13>> OH, MAN.
1:40:17>> OH, it's true. Cool. Yeah, that's a
1:40:20that's a good point. Um and so I guess
1:40:22like
1:40:23if if I if I am this the CEO of Mythical
1:40:26who's been acquired by Entropic, but I
1:40:27imagine he's still like working on it
1:40:28day-by-day.
1:40:30Um what would be your recommendations as
1:40:32someone who has successfully scaled a
1:40:34company to a much larger head count? Um
1:40:37what would be some of your
1:40:37recommendations for like what you would
1:40:39set up early to help with that
1:40:41transition with the understanding
1:40:43everything's going to get thrown out,
1:40:45but there might be some things I do
1:40:46differently.
1:40:47>> Yeah, I mean like I I try to think about
1:40:49it a bit like building a house, right?
1:40:50You wouldn't build a um a house without
1:40:53having a vision of what the house will
1:40:54look like in the end because the
1:40:55foundations you're going to lay would be
1:40:57rubbish, right? If you If you have no
1:40:58idea what the size and scale of this
1:41:00thing is that you're going to build,
1:41:01then you're not going to put the right
1:41:02things in place to be able to build it
1:41:03upon it. So, um yeah, if your company is
1:41:06always you Let's say you're going to um
1:41:08be a tiny therapeutics company that
1:41:09outsources everything. Like you're the
1:41:11foundations you need to lay internally
1:41:13within your organization can be
1:41:14relatively small, relatively modest.
1:41:16You're not going to need massive complex
1:41:17data management systems perhaps. But if
1:41:19you're planning to grow it to be a
1:41:20thousand person company that generates
1:41:23CRISPR modified cell lines at a rate of
1:41:25400 a day, like that's a different
1:41:28story. So, I think you you've got to
1:41:30you've got to think about the end goal.
1:41:32The way I try to think about it is think
1:41:33about the end goal, work back, and lay
1:41:35the foundations accordingly. Um
1:41:37obviously you can never think through
1:41:38everything. There are always going to be
1:41:39things that deviate or plans that
1:41:41change, but that's the way I try to
1:41:43contextualize it. It's think, okay,
1:41:44well, what's actually important for
1:41:46scaling from where we are now to where
1:41:48we want to get to as a vision. Um
1:41:50because otherwise, you know, conversely,
1:41:52you can do it the other way. You can
1:41:53start spending all your time putting
1:41:54these foundations in place and
1:41:55fundamentally you don't need them
1:41:56because actually where you're heading as
1:41:58a business and you're where you're
1:41:59actually planning to get to, you don't
1:42:00need those things. So, it it it cuts
1:42:02both ways, I think.
1:42:04>> Yeah, fair enough. Yeah, I've I've uh I
1:42:07I
1:42:08I wonder I wonder to what extent how
1:42:10that how that does change in the agentic
1:42:13era if you have like
1:42:15if you've got some of these different
1:42:17established agents and they are kind of
1:42:18working in lockstep with you to figure
1:42:21out how to shift the the blueprint of
1:42:25the larger house that you're building. I
1:42:26wonder how that changes going into the
1:42:28future. That's kind of like something
1:42:29I'm going through right now. I don't
1:42:30know if you've played around with Grok
1:42:32bot at all or anything or any of those
1:42:33tools yet, Ryan.
1:42:35>> Not that one. We use some of the other
1:42:36ones a lot, but no, not Grok.
1:42:38>> Yeah,
1:42:39well, I mean like more even like the the
1:42:41agentic side of it like with the with
1:42:43the individual bots and then I guess
1:42:45with Open AI it's bots and do that is
1:42:48just coming up with Anthropic.
1:42:50It it it's it's interesting to kind of
1:42:52see how that could be mapped on top to
1:42:55kind of help with that transition.
1:42:57>> And I think it also gives you an
1:42:58opportunity to kind of reset as well,
1:43:00you know, if you could you forgive
1:43:02forgive the comment, you could lay down
1:43:04a load of rubbish and then realize you
1:43:06need to sort it out later and AI is
1:43:07great at finding those patterns and
1:43:09sorting it out. So, I'm not saying you
1:43:10should do that as a strategy, but it
1:43:12gives you a kind of get out of jail free
1:43:14card in that regard. You could kind of
1:43:15say, you know, we've got all this stuff
1:43:17we've accrued over the last 2 years when
1:43:18the company was five people. Now we need
1:43:21to become 50 to 100 people. Sort that
1:43:23out.
1:43:24>> [laughter]
1:43:25>> And there was no tool to do that before
1:43:26on some very clever administrative
1:43:28assistant that would come down and sit
1:43:30through and sort everything out. So, it
1:43:32definitely does change the game in terms
1:43:34of what you can get away with and then
1:43:36make good, I think.
1:43:38>> Fair enough.
1:43:39>> Excellent. Okay. Um
1:43:41So, I guess like one one final one one
1:43:44final thought here.
1:43:45When biotech hands a program over to a
1:43:48development manufacturing partner, what
1:43:50would you say is kind of like the
1:43:51hardest thing historically that's that's
1:43:53been difficult to transfer, right? I've
1:43:54got a I've got a process inside of my
1:43:56lab. I'm interested in exporting this
1:43:58out. Um
1:44:01what do you kind of see like following
1:44:02falling through the cracks and something
1:44:04that could be short up potentially?
1:44:06>> Yeah, I'd say um
1:44:08so, there's there's obviously the
1:44:09process, the details of what they want
1:44:11these people, you know, that you're
1:44:12contracting to to do that work. Um
1:44:15that's kind of one side of it um that
1:44:17will hopefully mean that the
1:44:18relationship is productive between you
1:44:19and and that company. But, the probably
1:44:21the bigger issue is that chain of
1:44:23custody from where did your materials
1:44:26come from that they're receiving and
1:44:27then is pushing forward towards the
1:44:29clinic. Like that chain of custody of
1:44:31understanding how it was made. Like it
1:44:33could be as simple as a lost batch
1:44:34number, right? On a on a cell line that
1:44:36you were making that was going into make
1:44:37a vaccine or make a gene therapy or
1:44:39something like that. You lose that link,
1:44:41you lose that chain. Um when you're
1:44:42submitting it to the FDA or you know, or
1:44:45the European equivalent, like that's a
1:44:47problem. Like it's a broken chain in the
1:44:48in the link. And it can even be as
1:44:50simple as where did you get the stuff
1:44:51from initially? Did it come from a
1:44:52professor in South Korea that sent it to
1:44:54your PI back in 2001? Like it's you've
1:44:58got to kind of understand what the chain
1:45:00of custody is within what you've done,
1:45:01but also where did it come from before?
1:45:02So, that's kind of where a lot of people
1:45:04get caught out. I think my previous
1:45:05company we came across that quite a lot.
1:45:08Uh part of this inspiration behind Lab
1:45:10Thread is to try and help solve that
1:45:11like this, you know, the ability to
1:45:13track and log and understand how things
1:45:15chain through. Um but yeah, obviously
1:45:17you've got to pass the process over.
1:45:18Having a close working relationship with
1:45:20them as well. Like that that that makes
1:45:22a big difference. We've actually sent
1:45:23scientists over to some of our
1:45:25contractors and said like just go and
1:45:27show them how to do this um because, you
1:45:29know, that could that adds a huge amount
1:45:31of value in just getting it right first
1:45:33time. So, there are things you can do to
1:45:35mitigate it, but yeah, the processes
1:45:36themselves, but also ensuring that what
1:45:38you're passing on
1:45:40has the provenance you need to take it
1:45:41forward.
1:45:42>> Excellent.
1:45:44Forgive me if this is a dumb question,
1:45:46but Lab Thread, this is I'm now just
1:45:48realizing this. It's probably like the
1:45:50reason you built the thing.
1:45:52It would be like massively beneficial
1:45:54for technology transfer.
1:45:56>> Yeah, that's kind of you know, that's
1:45:58that's not the only point, but it's the
1:46:00thread in the name, right? It's like
1:46:02okay, so this this method of recording
1:46:04experiment in this electronic lab
1:46:06notebook is linked to the sample. This
1:46:07sample is linked to this piece of data.
1:46:09That piece of data is linked to the
1:46:10original design that validates the
1:46:12sequence is correct, whatever it might
1:46:13be. It's kind of that whole connected
1:46:15together, all of those things that gets
1:46:17built while you're recording it. So,
1:46:19that you know, that was the That was the
1:46:20pain I had at Oxygen like just trying to
1:46:23understand, you know, we had projects
1:46:24that would run over 2 years for cell
1:46:26line engineering. And we had to have
1:46:27every batch record recorded um in doing
1:46:30that. And you know, the team did a
1:46:31fantastic job of doing it, but it was a
1:46:32huge amount of work. So, yeah, that was
1:46:34a lot of the driver behind creating it.
1:46:36>> Mhm, that makes sense.
1:46:38Very good. Um I guess one one one final
1:46:40question we'll leave you on uh is just
1:46:42as well like what would be one piece of
1:46:44conventional wisdom uh in building out a
1:46:47biotech company that you've changed your
1:46:49mind on?
1:46:51>> Uh that I've changed my mind on. What
1:46:53have I changed my mind on? Why would Why
1:46:55would you do it to yourself? That's the
1:46:56main
1:46:57>> [laughter]
1:47:00>> I don't know. I mean
1:47:01>> normal job?
1:47:03>> [laughter]
1:47:03>> What am I doing?
1:47:04>> What are you
1:47:05>> What are you thinking?
1:47:07Yeah, um
1:47:09what have I changed my mind on? I I mean
1:47:10I I think that the main thing I realized
1:47:12is that I can convince myself of almost
1:47:14anything. Like it's um
1:47:16you can really tell yourself that
1:47:18everybody's going to love this product.
1:47:19It's going to be the best thing in the
1:47:20world. They just need to learn If I just
1:47:21knew about it, then I'd have a a billion
1:47:23in sales.
1:47:24You can really convince yourself of
1:47:26pretty much anything. And then about 6
1:47:28months later you're like, "Oh, that's
1:47:29why they weren't buying it. Nobody
1:47:31[laughter] needs it."
1:47:32>> Sure.
1:47:34>> So, yeah. I wouldn't say I've
1:47:35necessarily changed my mind on it, but
1:47:37I've realized the limitations and
1:47:39fallibility of your own human
1:47:40conviction.
1:47:41>> Right.
1:47:42>> But
1:47:42>> Yeah. So, basically it would be that it
1:47:44would it would be one of those key
1:47:45pieces of information which is basically
1:47:47like you got to go talk to your
1:47:48customers. Um we can you can psych
1:47:50yourself up with regards to what you
1:47:52think is the the market wants, but you
1:47:54need to make sure you actually go out
1:47:55and talk to people about it first.
1:47:57>> And so, I it's quite scary but doing
1:47:59that because particularly let's say you
1:48:00worked on a project for like 4 years,
1:48:01right? It's your research. It's your
1:48:02baby. And then you're going to send take
1:48:04it to somebody and they're going to say,
1:48:05"Oh, this is rubbish." And
1:48:06>> This is stupid. Yeah.
1:48:07>> Yeah, exactly. That's like the worst
1:48:09thing for you, particularly when you're
1:48:11young as well. Um now I'm a bit older,
1:48:13I'm like, "Just What do you guys think?
1:48:14Do you like it? Does it work? Does it
1:48:16not? Like, you know, do you like the
1:48:17color? Whatever." Like, it's I'm fine
1:48:19with it now, but you know, as you get
1:48:20older you have thicker skin and stuff
1:48:22like that. But, it's difficult when
1:48:23you're young. If you think it's going to
1:48:24fly and it's going to be the next best
1:48:25thing, and then you don't want to you
1:48:27don't want to show it to somebody else
1:48:28in case they say it's not. That's
1:48:29actually quite a risk when you're
1:48:30running a company.
1:48:32>> He's talking to us, Ben.
1:48:34>> [laughter]
1:48:35>> And he doesn't even know it.
1:48:38>> It's going great, guys. It's going
1:48:40great.
1:48:40>> [laughter]
1:48:40>> Yeah.
1:48:41Excellent. Right on. Well, Ryan, thank
1:48:44you so much for your time. We appreciate
1:48:45you coming on. We're we're appreciative
1:48:47of your support, and we're looking
1:48:48forward to seeing uh uh more and more
1:48:50scientists using something like LabThred
1:48:52to make sure that uh they don't get into
1:48:54that headache of chain of custody, for
1:48:56sure.
1:48:56>> Indeed. Indeed. All right, guys. Take
1:48:58care. Bye.
1:48:58>> Thanks, Ryan. Thanks so much for coming.
1:49:03All right.
1:49:04>> All right.
1:49:05>> Excellent. So, uh
1:49:06Jeff, uh just quickly here at the end,
1:49:08one of the things that I wanted to spend
1:49:10a little bit of time on um would be
1:49:12going through uh
1:49:15the
1:49:16the the timeline, just to see if there's
1:49:18a a couple other things that we wanted
1:49:19to to pull up, if there's anything that
1:49:22we wanted to spend some time on. Um
1:49:24>> The timeline.
1:49:25>> The timeline. I figured I'd just open up
1:49:28X and see if there's anything that uh
1:49:30that [clears throat] that would be of
1:49:31that would be of interest for us to look
1:49:32at. Um
1:49:34>> I'd be careful opening up X.
1:49:36>> [laughter]
1:49:37>> Just
1:49:38my raw algorithm.
1:49:39>> So, no, actually no. So, actually one of
1:49:42the things I wanted to pull up in
1:49:43particular that we could spend a little
1:49:44bit of time on um is
1:49:47um oh, okay. Perfect. It's already
1:49:50shared. Um
1:49:52One of the things I wanted to spend a
1:49:53little bit of time on is actually I saw
1:49:54a really interesting post from from
1:49:56Jason Kelly
1:49:58um over at Ginkgo Bioworks. He was
1:50:00talking about Mithril. Let's see if I
1:50:01can find it. Oh, yeah.
1:50:03Um
1:50:04where he basically I don't know if I can
1:50:06maybe I can I don't know if I can zoom
1:50:08in on this at all or not. Um let's see.
1:50:11Ooh, there we go. All right. Perfect. Um
1:50:14so, he was he so he's basically
1:50:15responding to the to the uh to to the
1:50:19news, right? That Mithril was launched.
1:50:22And one of the things he talked about
1:50:23was awesome to see this, really excited
1:50:25about the one person biotech in
1:50:27particular. CROs with APIs available for
1:50:31agents like Ginkgo's data point service
1:50:33plus something like cursor for biotech
1:50:35could really change the cost of
1:50:37launching biotech startups.
1:50:39Like Jeff, this is kind of where I see
1:50:42this potentially going, right? Is
1:50:45could we get to a point
1:50:47>> That's actually crazy.
1:50:49>> Yeah, where you've got Claude and you've
1:50:51got your your agents just directly
1:50:54interfacing with your CRO that is hooked
1:50:57up to a bunch of autonomous bots.
1:50:59Um and you're you're basically just
1:51:02paying for time inside of the inside of
1:51:04the the the physical world of the
1:51:07>> [sighs]
1:51:07>> lab automation bots.
1:51:08>> What is going on?
1:51:11Isn't that wild?
1:51:12>> Isn't that wild?
1:51:14>> I didn't even think of like a like a CRO
1:51:16API.
1:51:17>> Yeah.
1:51:18>> That's huge.
1:51:20>> Yeah.
1:51:20>> Dang.
1:51:21>> Ginkgo's a really cool company, man.
1:51:24I've been I've been for a tour of their
1:51:26labs a couple different times. This was
1:51:27years ago and they've like updated since
1:51:29then and just blows your mind. It is
1:51:32like something out of Star Wars. They've
1:51:34got like these
1:51:35>> dude, we should try and start a biotech
1:51:36company.
1:51:37We should just do it. We I mean like
1:51:39what are we doing here?
1:51:40>> [laughter]
1:51:40>> That should be what we do. That'd be so
1:51:42awesome.
1:51:44>> It'd be It'd be super interesting, man,
1:51:46to basically document our journey in
1:51:48like the last 15 minutes of the biotech
1:51:50forager every week and just be like,
1:51:52"Hey, here's what we've been doing.
1:51:53Here's what we've been working on.
1:51:54Here's what's worked. Here's what
1:51:55hasn't. Here's what we've
1:51:58Like I don't think this has ever been
1:51:59done before. Could we actually map out
1:52:02and just document our journey?
1:52:05>> I don't know if that would be like we're
1:52:07exposing proprietary information. Like
1:52:09you we could we probably couldn't do
1:52:10that.
1:52:11>> We probably We could probably figure out
1:52:12a way to do it.
1:52:15>> Figuring out while live on live stream.
1:52:18>> [laughter]
1:52:18>> Well, no. Like we wouldn't we wouldn't
1:52:19do like we wouldn't review the data.
1:52:20We'd just give people updates on a on a
1:52:22weekly basis.
1:52:23>> Ah, got it. Yeah, we can do that.
1:52:24>> Yeah.
1:52:25>> I mean, that Yeah, we should look into
1:52:26that because that would be that I mean,
1:52:28like if it's getting that that changes
1:52:30my mind a little bit.
1:52:31If this if CROs have APIs available, I'm
1:52:33wondering if you got to pay for it.
1:52:35>> Of course. Well, you're not going to be
1:52:37able to work for the CRO for for free.
1:52:39The CRO
1:52:39>> no, no. I mean, like the API. I feel
1:52:42like if they offer you API for free and
1:52:44then it's like, "Yo, we have to use the
1:52:45CRO now because we have all of our plans
1:52:48through their API, right?"
1:52:51Um [snorts]
1:52:52dang. Okay, I'm going to bookmark that.
1:52:55Ginkgo's model API. Geez.
1:52:58>> Yeah.
1:52:59>> I'm actually Let me read through this
1:53:00real quick because today Okay, so this
1:53:01was from September 16th, 2024, so about
1:53:042 years ago. Um today we're thrilled to
1:53:07announce a bold new step for Ginkgo
1:53:08Bioworks, the launch of our model API, a
1:53:11powerful tool designed to bring
1:53:12biological AI models directly to machine
1:53:14learning scientists. Powered by our
1:53:16partnership with Google Cloud, we'll be
1:53:18making this API publicly available on
1:53:20our website today. Dude, it's free.
1:53:22Researchers, developers, and enterprise
1:53:23teams will also be able to access the
1:53:25models that power our API on Google's
1:53:26Model Garden soon.
1:53:29With this programmer-friendly,
1:53:30ultra-low-cost API, Ginkgo is making our
1:53:32internally developed AI tools available
1:53:33to anyone, and we couldn't be more
1:53:35excited to begin sharing our work. The
1:53:37interface provides an easy and scalable
1:53:39way to access sophisticated models
1:53:40trained on protein and DNA data,
1:53:42starting with our first release,
1:53:44{colon} a machine learning model trained
1:53:47on proprietary Ginkgo datasets.
1:53:50Read more about our first model, AA0, a
1:53:53large-scale model trained on 2+ billion
1:53:55proprietary Ginkgo protein sequences
1:53:57here. Oh my gosh.
1:54:00Dude,
1:54:01we should just do it.
1:54:03We should just do it.
1:54:06All right.
1:54:07>> What modality, Jeff?
1:54:08>> BioTech Voyager. Jeff and Ben start a
1:54:10biotech company with Ginkgo Bioworks.
1:54:12>> [laughter]
1:54:13>> And Vlad.
1:54:14>> What would what would what modality
1:54:16would you want to do?
1:54:17>> I'm not going to tell. We can discuss
1:54:19this after we get after we stop. After
1:54:22we get off the live stream.
1:54:23>> Fair enough.
1:54:25All right. Well, folks, we're going to
1:54:26get off the live stream.
1:54:27>> [laughter]
1:54:30>> But actually, we're probably going to
1:54:31get off the live stream. But thank you
1:54:32so much for joining. Uh
1:54:33it's been a it's been a I think it's
1:54:35been a it's been a while, Jeff, since we
1:54:37gotten close to the 2-hour mark. We're
1:54:38we're getting closer. We're we've got a
1:54:40lot of exciting guests lined up going
1:54:42into the future. So, expect to see more
1:54:45of us
1:54:46uh and more and more more content
1:54:48coming, too.
1:54:49Excellent. All right. Well, until next
1:54:51time, folks. Uh thank you for joining us
1:54:53on the Biotech Voyager. And uh we'll see
1:54:56you next time. Cheers.

About this transcript

This page contains the full transcript of Anthropic Acquires a 1-Person Therapeutic Company, ARPHA-H GIVE Program Highlight, and Much More... by The Biotech Voyager, generated from the public captions YouTube serves with the video. The transcript has 20,842 words across 3,512 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.