Episode 8: From Bench to Boardroom: What Drug Development Actually Takes

[00:00:00] 

William: I always have what's called I call the grandma test.

Armon: Yeah. 

William: Which is like what you put your own grandmother on this trial. 

Armon: Yeah. 

William: And if you wouldn't, then at least for me, then you really shouldn't be doing 

Armon: Totally agree 

William: developing that molecule. So I would just keep that standard. For myself at least. 

Armon: A hundred percent agree. 

William: Okay. ​ 

Armon: normally I ask people to give a background on themselves, but yours is too long. Uh, so we'll just hit it as we go.

Okay. Well, what's your one minute version of your background? Uh, let's see. I'm a physician scientist that started out in academia and went to pharma and is now in biotech. Yeah. 

So what made you wanna leave? I mean, you were running a lot of cool stuff at Vanderbilt and then you decided to go to the pharma side at Roche.

What, what made you wanna make that jump to the pharma side? 

William: Yeah, so in my academic career, so I'm a physician scientist by training. I saw patients had a lab, helped to run trials, uh. [00:01:00] During that academic career, I was focused mainly on targeted therapies in lung cancer. 

Armon: Yeah. 

William: And I also helped other companies develop their molecules.

Yeah. So for example, I helped AstraZeneca develop Azd 9 2 9 1. Yeah. Which became, uh, Osier Number Tgso also helped a small company called X Discovery, develop another molecule, which actually was also approved in Sartin. 

Armon: Yeah. 

William: Um, and then I got headhunted, uh, and. Most of the headhunting offers I had turned down.

Armon: Yeah. 

William: But this one was from Roche, and Roche is very scientifically strong. 

Armon: Yeah. 

William: Uh, particularly on oncology, it was to head up oncology and, uh, I did agonize for a long time, but you know, after I met the folks there and weighed the pros and cons and thought about what I would do in the future and academia versus in industry, I decided to make the plunge.

And the main reason was that I wanted to be able to do drug development full time. 

Armon: Yeah. 

William: And then ultimately have impact on patients. Yeah. Uh, you know, maybe hundreds, thousands, maybe even more than [00:02:00] that. Uh, whereas I do miss seeing patients one at a time. 

Armon: Yeah. 

William: But the impact from drug development, it can be much greater.

Armon: So it's like a trade off of, you know, you have a whatever, a hundred percent chance of helping a few thousand a year versus like, I have a 20% chance now of helping millions. 

William: I dunno if it's 20%, it could, well, 

Armon: you know what I mean? 

William: Yeah. 

Armon: I'm saying with your selections it was higher than average. Yes. 

William: Yes. My PTSDs, yeah.

Probability. 

Armon: What, what do you think, uh, was most different about pharma relative to what you expected when you were going in. 

William: Yeah. So, um, you know, I found that the science I think is equally the same quality, like the quality of the scientists and the stuff that's being done. Yeah. But obviously in pharma it's much more directed.

Armon: Yeah. 

William: Whereas in academia, you can really follow any idea what you want. 

Armon: Yeah. 

William: Um, and you have more freedom to do that. Um, the other main. Surprise was the number of people in pharma that are experts in, uh, diverse areas. 

Armon: Mm-hmm. 

William: Yeah. So, um, [00:03:00] as an academic I usually interacted with one or two people in the pharma company, like the key scientist on the project or the key clinical person.

Yeah. And then I got to Roche in Basel, Switzerland, and I was like, wait, there's an expert for formulation, there's an expert for, um, you know, uh, whatever topic, toxicology, uh. Uh, dosing and so on and so forth, and just, um, sort of realizing that, wow, there's all these scientists that really do. Amazing work, but in particular areas and they all contribute to development of the molecule.

Armon: Yeah. 

William: It's sort of like a tertiary can, um, you know, a tertiary hospital. 

Armon: Yeah. 

William: Versus a primary care hospital. Yeah. Yeah. And in a tertiary care hospital, you have just levels of expertise. 

Armon: Well, well, I think there's also like, I mean, I was coming at it from a student angle, but there, I think there was this misconception that.

Oh, we, the academics are the smarter ones, and then the pharma and biotech people are slightly stupider on average. And then you actually go there and you're like, you know, they're actually pretty damn good. 

William: Yeah. I mean, I think growing, growing up when I was a fellow, I probably [00:04:00] had that, uh, attitude. 

Armon: Yeah.

William: But I think, um, you know, in today's world, you can't be a successful pharma company without having. Scientist. 

Armon: Oh yeah, yeah. No, it was, it was very, I think it was just like one of those weird misconceptions that the academic side would nurture for some reason. 

William: Yeah. Well, I mean, well like in oncology, you know, in the, not with no disrespect.

Armon: Yeah. 

William: But in the eighties and nineties, a lot of it was about developing chemotherapies, right? 

Armon: Yeah. 

William: And then a lot of it was empiric. 

Armon: Right? 

William: That's true. Like a mechanism based thing. Yeah. So I think some of it came from that baggage. 

Armon: At least 

William: in oncology. 

Armon: Yeah. Yeah. 'cause they're like, we basically have these poisons, let's mix them in cocktails and see what happens.

William: Yeah, yeah, yeah. And I even know the cancer biologist would look down on the clinicians and be like, you just are giving poisons. Whereas we're figuring out the mechanism. But now, 

Armon: yeah. Yeah. 

William: With targeted therapy and you'd have to totally understand the mechanism. So, 

Armon: yeah, that makes sense. 

Armon: so what does the job actually entail when you're on the pharma side?

So like, let's start with head of oncology. What's the day like? What are you up to? [00:05:00] Especially what are the things that you think are non-obvious to people? 

William: What are non-obvious? Yeah. Well, I mean, every day's totally different. 

Armon: Yeah. 

William: Um, you can be interacting. Well, I was the head of oncology Yeah. For Roche for four years.

Uh, the Roche Pharma research early development side, and then I was head of whole period. Yeah. So in oncology, yeah. I mean, you can be interacting with a scientist one day. You have to be interacting with the clinicians the other day. You're looking at data, whether it's clinical or basic science. 

Armon: Yeah. 

William: Um, you have to figure out what the strategy is.

Yeah. Um, you know, I was fortunate to hire folks like Christian Rahmel 

Armon: mm-hmm. 

William: And Friedrich Stein, uh, as my heads of discovery and clinical respectively. Yeah. Um, and we had to, you know, I think one of the things you have to do in drug development is envision the future. Yeah. And not only envision the future, but like figure out like, if I'm working on something today, is it still gonna be relevant by the time?

It gets approved in say, five to 10 years. 

Armon: Yeah. 

William: Five if you're lucky, right? Yeah. Probably more like 10. Um, so a lot of that strategy is really important. 

Um, [00:06:00] 

Armon: and, and what's the feedback loop like between kind of your vantage point versus let's say pharma's obviously an organization with many stakeholders like the commercial side?

William: Yeah. 

Armon: So you're like, oh, I think if we wanna have maximum impact, let's do X. And then commercial brings in a perspective like, oh, actually we think y. How does that work there? 

William: Yeah, so I was fortunate that, um, my, uh, CEO at the time was sever schwan. 

Armon: Yeah. 

William: Who would actually, and this is not a secret 'cause he would say it in public too.

He would tell the research guys not to listen to the commercial guys. Okay. 

Armon: Which is different from the other pharma you feel. 

William: Well, let me get back to that. Yeah. Okay. So his point was that he, um, he actually had a famous, uh, laminated slide in his office, 

Armon: Uhhuh, 

William: and would say like, well, look on the top of these are the, um, probability of technical success of the molecules in development.

On the bottom are the loss of exclusivity. 

Armon: Mm-hmm. 

William: And 90% of the data [00:07:00] on this. One slide that I've laminated is wrong. And his point was that like, you know, Mapthera, um, which is a CD 20 antibody thoma, you know, was originally projected to have like 200 millions of sales and it became a, I think a 7 billion blockbuster.

Yeah. Um, conversely there were other medicines that were predicted to have billions of sales that didn't even make it. 

Armon: Yeah. Yeah. 

William: So I think his point was very early on in the life of a molecule. It's hard to tell how big of an impact it's going to have. 

Armon: Yeah. 

William: Now there's some grains of salt with that, which is like if you're going after a genetically defined disease, in a rare disease population, that's only one in a million.

Armon: Yeah. 

William: You know, you might know that it's only gonna be one in a million. 

Armon: Right, right. 

William: But if you're developing something like CD 20 for the first time. You know, anti CD 20 antibody. It might work in one area, but then you don't know that it's eventually gonna work in say, 10 different areas or something like that.

Yeah. So I think his point was follow the science. At some point you do have to make a commercial, um, decision. 

Armon: [00:08:00] Yeah. 

William: Because in a big pharma company, a molecule should bring in enough revenue. Right, 

Armon: right. 

William: Get into the math about that if you want. Yeah. But early on, hopefully you're, you know, just do the best science and follow that.

Armon: Yeah. 

Well, actually a little bit of the math might be helpful for folks just to get a sense for like, what is this really tough equation people are trying to balance? Because I, I think from the outside at superficial, they're like, oh, it's 90% margin on these drugs. Like you're selling it for 200 k, it costs you sub 10 K to make.

Um, like per cogs, but they don't understand all the failure rates and everything else that goes on. So what's your perspective on the map there? 

William: Yeah. Okay. Uh, and then I want to come back to the, from commercial thing too. Yeah. Well, first I'm finishing on the commercial. 

Armon: Yeah. 

William: There are now shifts for some companies saying like, you can't even work on a molecule and discovery unless it's gonna have more than a billion dollar potential.

Armon: Mm-hmm. 

William: As if you could predict that at a, at a very early stage. So anyway, so to come back to the math Yeah. So. So, yeah. Well, 

Armon: sorry, because you got me going now. [00:09:00] 

William: Okay. 

Armon: Do you think like, and maybe this is the wrong perspective, part of me has always thought of like. In pharma, commercial is closest to the money, let's call it.

Whereas let's say in biotech, it's actually research that close because the money is the investor money and it's based on r and d data. No one's actually selling anything. 

William: Mm-hmm. 

Armon: So like r and D can drive it more, whereas commercial side, it's like. If they spent a hundred million dollars and increased sales by 10% across the board, that is worth it all day, every day.

Mm-hmm. Versus like taking a punt on another drug. So is that where that influence comes from in a way that they're most proximal to what the investors and the headlines care about? Uh, whereas in biotech, it's actually r and d that's most proximal to it. 

William: Well, here's the way I think about it from a math perspective.

So I had $1.2 billion a year of budget. 

Armon: Yeah. 

William: And on average over 10 years. Over five years, I was supposed to hand over 10 molecules to late stage. 

Armon: Yeah. 

William: So if you do the math on that, that was when I was, yeah. Yeah. On calling. If you do the math, then it's two molecules [00:10:00] per year on average. 

Armon: Yep. 

William: 1.2 divided by two is 600 million.

Armon: Yeah. 

William: So I'm spending basically 600 million per molecule to hand it over to late stage. 

Armon: Yep. 

William: And then when late stage takes it over, they need to probably spend maybe a hundred, 200, at least million two. Do the phase three. 

Armon: Yeah. 

William: And maybe sometimes more. So by the time that molecule's approved, you spent like a billion dollars.

Yeah. It sounds crazy, but that's what it is. Yeah. But the 600 million per molecule that you hand over also takes into account all the other molecules that potentially didn't make it. 

Armon: Yeah. 

William: Now of course, if we had a higher success rate, then you know the number would come down, but on average, that's the way it is.

So basically, if I'm spending 600 million per molecule, 

Armon: yeah. 

William: My molecule eventually when it's commercialized, doesn't make 600 million. I'm losing money for the company. 

Armon: Yep. 

William: And so over 10 years, I would've lost. Yeah. A lot of money. Right? Yeah. So that's sort of where the, I think the blockbuster number comes from.

Armon: And what do you think is the greatest source of [00:11:00] inefficiency? 'cause obviously there's the attrition rate of any molecule going in, but why is it that it's, you know, for 1.2 billion you're only gonna get two molecules, whereas it's not four. Like what would it take efficiency wise? 

William: Yeah. I mean, I think if you only worked on validated targets 

Armon: Yeah.

William: And you knew all the biology and you knew exactly which patient populations to go into, obviously you could have a much higher 

Armon: Yeah. 

William: What we call PTS probability of technical success. 

Armon: Yeah. 

William: But you'd probably also only be making me too, me better, maybe even me worse molecules. Yep. So I think you have to have a mix of what you think you can improve on.

Yeah. That are established targets. And then you do have to work on some novel targets. Now you have a portfolio in pharma. 

Armon: Right. 

William: So I don't think you'd work on totally all novel targets because you might produce nothing. 

Armon: Yeah. 

William: Um, but, you know, it's, it's really interesting. Um. As you probably know, every molecule that ever gets approved has to get to this intersection of biological understanding.

Yep. Clinical understanding of where the molecule might be relevant, and [00:12:00] then technological advancement to make that target drugable. 

Armon: Yep. 

William: And getting to that sweet spot, the Venn diagram Yeah. Uh, is not that easy. 

Armon: Right. 

William: Uh, particularly for new targets. 

Armon: Yep. 

William: Um, and so for pharma, if you really want to, well, and even biotech if you want, um, to be successful in a short period of time.

It's much easier to go after a validated target. 

Armon: Yeah. 

William: Uh, than a, a, a unproven target. 

Armon: Do you think I, I know this is like somewhat tangential, but the incentive for making me too molecules, how much of it, or what do you think. Would remove more of that incentive. Because my impression is that it's mostly 'cause especially us commercial, you can have many drugs that are otherwise basically the same thing.

William: Yeah. 

Armon: Coexist and they will be commercially worth it. Whereas in other jurisdictions where someone just decides, Hey, we're just buying this one and the rest don't have a play, that really kills the incentive for me too drugs. So is that what drives it or what do you think drives it? And if we got rid of the Me Too incentive [00:13:00] somehow, would it actually bias us towards a lot of new.

You know, new generation molecules. 

William: Yeah, that's a tough one because the challenge on the novel targets is, as I mentioned, it can take a long time. 

Armon: Right? 

William: Right. Uh, and so if your horizon is 20, 30 years, or you have a long, you know, investors that are in it for the long haul, then I think you'll be okay. Yeah.

But if you have short term. Investors where if you're in a pharma company that has a patent cliff coming Yep. And you need to generate that revenue quickly, otherwise your stock will drop and so on and so forth, then you do have to go after these, um, probably more clinically validated or clinically validated can be that there's a molecule approved.

Armon: Mm-hmm. 

William: Already. Or it could be that your competitors working on something and then a phase one, like they showed like amazing data and then everyone jumps into that Van Viken. 

Armon: Yeah. Yeah. 

William: Now what is interesting is you've probably seen some of these analyses. Now there's more hurting. So to speak than at least in, uh, the last 20 years.

Yeah. I think people analyze the big pharma pipelines. 

Armon: Yeah. 

William: And there's a lot more crowding [00:14:00] now on Val, on the same targets rather than I think 20, 25 years ago. And 

Armon: because of the dynamics you mentioned that like nearer term, it's more likely yield to hit. 

William: Yeah. Yeah, 

Armon: yeah. 

William: Now of course, if it also, the market has to be big enough, right?

Yeah. So GLP one, we see that there's a ton, uh, but that market's expected to be 20 billion, 

Armon: right? Right. 

William: If unfortunately the commercial. Reality is that if it was a market of 200 million, I'm not sure you'd have so many people 

Armon: For sure. 

William: Getting into that space. 

Armon: So with all these dynamics that you were balancing, what was the most difficult part of the job to you?

William: Um, yeah, I mean, I think there's, it's, it's an amazing job, is really fun. There's a lot of privilege to be able to do that. I think, you know, you have to deal with a lot of, um, stakeholder management. 

Armon: Mm-hmm. 

William: In a big company, there's lots of different people. So even in your own organization, like you have to.

There's different functions. 

Armon: Yep. 

William: You know, there's the, uh, therapeutic area group, then there's, uh, therapeutic modality group, the pharmaceutical science [00:15:00] group. So getting everyone aligned, uh, on the same vision and so forth. Yeah. Uh, and then, and then transferring over to late stage, you have to get a whole bunch of other people up to speed and Yeah.

Convince, uh, you know, whether it's worth investing or not. And, um. And then, you know, and then having a large organization that you have a lot of, uh, human resource, uh, yeah. Topics to deal with 

Armon: as well. Of course, 

William: yeah. 

Armon: Now what are, what were the aspects of this that you saw that you feel are most different from what people expect?

Like even within the industry, everyone has their own like image of what someone like you does, but what, what's the aspect that you think is most different from what they think? It's, 

William: oh, that's an interesting question. I mean, I think the higher up you go. Uh, and more responsibility you have, maybe the less close you can be to the science.

Yeah. So that's one thing. Uh, let's see, what else? Um, I think the main thing, well, okay, so also going from academia to industry. 

Armon: Yeah. 

William: Yeah. I mean, academia in my mind is a lot more about [00:16:00] individual accomplishments and what have you done versus. Pharma and biotech was a lot more teamwork. 

Armon: Yeah. 

William: And the way I sum it up is, um, in academia, I probably had to update my CV like every week.

Armon: Yeah. 

William: You know, so it's like, what was the latest senior author 

Armon: on this one? 

William: Exactly. Yeah. And then, you know, who did I train? And, and so on and so forth. Or I have to submit my IH cv. Yeah. Uh, every day by 

Armon: sketch. Yeah. 

William: Bio sketch. Yeah. Whereas in pharma, I don't think I updated my CV for like four years. 

Armon: Yeah, yeah.

William: Yeah. 

Armon: No, I mean I definitely saw the team sport elements of it because I, because I think that's one thing, speaking of it from like the junior scientist side, that I think some scientists like end up hating academia for the like very individualistic, somewhat cutthroat at times, because like if you're first author versus second author, that can be everything in certain situations.

Yeah. And so it's very non-team oriented. And then like industry is like, all right, we're all gonna sink or swim together. 

William: Yeah, yeah, yeah. 

Armon: This molecule hits, everyone's gonna be happy. If it [00:17:00] doesn't hit, we're all screwed. 

Armon: you know, obviously with all these different stakeholders involved, there's a lot of political pressures.

By, I mean, like within a company, political pressures, how often did you feel like there were situations where there was political pressure to advance a drug that in your heart you're like, no, I really wouldn't bet on that one. Uh, or kill a drug that you're like, I, you know, this one's gonna be awesome.

This would be one of my two for the year, but it, it gets shot down. Like, how often was that? Problem. And how does one navigate it? 

William: Yeah. The shooting down one, there was a particular molecule, uh, that I was working on that, uh, I won't name it, but it was what molecule? It's a target I've worked on in the past.

Armon: Yeah. 

William: So you can guess. Yeah. Uh, but, uh, the, this mechanism of acquire resistance was coming out to existing therapies. Mm-hmm. 

Armon: Yeah. 

William: It wasn't clear when the initial data came out, how frequent that resistance mechanism was gonna be. 

Armon: Yeah. 

William: But by the time we got a small molecule to overcome [00:18:00] that mechanism of resistance 

Armon: mm-hmm.

William: It was a small fraction. Uh, and there, I would have to say, unfortunately, and I, it pains me to say this, but I did have to take a commercial. Consideration. 

Armon: Yeah. 

William: And say, you know, this is a great, well, the molecule had some liabilities also, but given that and then the tiny patient population, 

Armon: yeah. 

William: It was hard to justify.

Armon: Yeah. 

William: And I really. Felt bad about it. Yeah. You know, because in academia I would never do that. 

Armon: Yeah, yeah, yeah. 

William: Um, but I had to do that 'cause it was right for the company. 

Armon: And, and are there constructive enough mechanisms for these things to find another way to live on, or do you think feel like they're just like gone 

William: shelf?

Oh yeah, no. So those kinds of molecules can be licensed out and then, um, you know, there's different size companies that are fine with different levels of, um, revenue. So a big pharma company will need obviously a lot of revenue from its molecules, but there are mid-size or smaller size that are fine and happy with less revenue.

So I do think in the long run, [00:19:00] those molecules will make it. 

Armon: Yeah. 

William: Um, and so hopefully in today's ecosystem that's possible. 

Armon: Right. 

William: The other thing that I also sometimes got frustrated with. Because I'm a big proponent of personalized medicine. 

Armon: Yeah. 

William: Particularly in cancer. Mm-hmm. And some of the mutations get very, very small.

Armon: Yeah. 

William: Uh, of fre, you know, the frequencies. Yeah. Yeah. And the issue is, um, you know, during phase one and phase two, a lot of academics want you to attest their, your molecule in these specific small. Uh, subgroups. 

Armon: Yeah. 

William: But it becomes challenging, right? Because you need the biggest indication 

Armon: Yes. 

William: Initially for the late stage commercial guys to accept it so.

I'd have to hold off doing those studies. But then with the promise that once it's approved 

Armon: Yeah. 

William: You know, from a lifecycle management perspective, those studies can be done. So it's not like it won't ever happen. 

Armon: Yeah. 

William: But it just may not happen now that that was another painful thing from going from academia to industry.

Armon: And actually one other curiosity that I just thought of as, you [00:20:00] know, 'cause you were head of Prad and then like at Pfizer, you were Chief development officer. Was there a disease area that you felt the dynamics there was just way different from the other ones? I'm just curious because I'm at least very used to oncology, but.

William: Oh yeah. I mean, uh, I guess the way I used to think about it is what is like what therapeutic areas are, like precision oncology, 

Armon: uhhuh, 

William: you know, you have maybe a genetically defined patient population. Mm-hmm. You know exactly which patients you wanna go after. You have a therapy that hits that with a good therapeutic index and so on and so forth.

Meaning like the mechanisms well known. 

Armon: Yeah. 

William: And so that would be like, you know, like precision oncology. Would be infectious disease where you know, the, 

Armon: you know what virus you're hitting 

William: Exactly. 

Armon: Or bacteria you're hitting. 

William: Yeah. Um, or maybe some rare disease where again, you know, the genetic Yeah.

Situation. Then there's the flip side, which is you have no idea what, what the mechanism is. Yeah. Uh, and so on. And so you then you have to decide like, you know, how valuable is this? How, how worth [00:21:00] the, is this program of moving forward? 

Armon: Yeah, yeah. Because it's like so much more uncertain what's gonna happen later.

William: Yeah. Now, I do think having trained in oncology, as you know, you know, we have concepts of tumor initiation, progression, and maintenance. Yep. Um, and that's very helpful. It's like what genetic lesion or what happens to make the tumor start in the first place and what makes the tumor continue to grow. 

Armon: Yeah.

William: Spread, uh, spread. And then what, what is needed? You, the initial mutation that you had may not be needed later on. Right. Yeah. And these are all concepts that are really important for drug development. 

Armon: Yeah. 

William: And I do think some of the other therapeutic areas don't have the same level of. Yeah. Of understanding.

Yeah. With no disrespect to those areas. 

Armon: No, no, no. It makes sense. I think some of those just like might lend themselves less to it or it's too complicated to figure out. 

William: Yeah. Yeah. So from a portfolio perspective, then you sort of have to balance. 

Armon: Yeah. 'cause I could see like autoimmunity be being one of them that's like, some of them are so fricking hard to figure out what 

William: Yeah, 

Armon: yeah.

Many have tried and failed. [00:22:00] 

Like one interesting dynamic I think that you had to live with is this aspect of you, for example, as head of r and d, would make certain decisions that will not have readouts until way after you're probably gone. 

William: Yeah. 

Armon: Uh, and I. Also you're dealing with the consequences of decisions made before you 

William: Yeah.

Armon: That you had nothing to do with, but they like have their readout under your watch. What's that like? Like how do you deal with it? Because I could see it being like your fault, even though it wasn't 

William: Yeah. Well, 

Armon: or your greatest work won't be known until, 

William: unfortunately, if you're in the leadership position and happens under your watch, it's your problem.

Or you're, it's either you're, um, you're a hero or it's your challenge. 

Armon: Yeah. 

William: Um, but yeah, I think the best advice I got from. Someone, when I first went to Roche, actually the chief legal c at the 

Armon: time, Uhhuh 

William: was like, yeah, you're um, only as good as your predecessor at the be, you know, the first three to five years or so.

First three years. 

Armon: Yeah. 

William: And then after that sort of it's you, uh, you know, and your [00:23:00] decisions and stuff like that. So yeah, I think in pharma. Drug development, you have to realize that you're on a relay race. 

Armon: Yeah. 

William: You know, you'll be there, but you have to make the best decisions during the time that you're there.

Yeah. Hopefully you build a pipeline that will produce when you're gone. 

Armon: Yeah. 

William: Um, and then hopefully the person that was in front of you also did the same. 

Armon: Yeah. 

William: But as you know, probably, you know, when you take over as the new head. Often you do a review of the portfolio. 

Armon: Right. 

William: And many times you end up killing a bunch of stuff.

Yeah. Uh, that, uh, is unlikely to succeed. And 

Armon: the prior, the prior heads like, God dammit, that was my pet project. Yeah. 

William: I mean, a lot of times it's obvious what you should stop. Um, and coming back to, you know, what's hard is like Yeah. I mean, I think, um, a lot of teams do get invested in their programs. 

Armon: Yeah.

William: Um, they, some might have spent even five to 10 years on them and so can become biased. Right. Yeah. And so, um, somebody has to take that decision. And I realized actually as the [00:24:00] head of the group, that even my therapeutic area heads who I thought were as unbiased as possible, even they had to like, represent their group.

Armon: Yeah, yeah. 

William: So on and so forth. So actually it was part of my job to be killing programs. 

Armon: Yeah. 

William: I mean, if I didn't kill 'em, who else was gonna do it? Yeah, yeah. 

Armon: No, you were just like gonna drown. 

William: Yeah, yeah, yeah, yeah. And killing them, you know, frees up, um. The re, the source, the resources, and the people that work on stuff that's more likely to work.

Armon: Yeah. And maybe a weird question, like, I don't know A how much contact you would typically have with other heads of r and d, but I'm curious across, because I mean whatever we're dominated by 10, 20 pharmas total, how often do you think the heads of r and d actually agree with each other, but the.

Portfolios looking different is due to other forces and other dynamics or versus like, these people actually have very different views on what should be done. 

William: Oh. I mean, I think when you go to a place that you have to really deal, uh, play to the strengths, right? Yeah. You can't just go to a place and say like, I want, um, [00:25:00] the other company's portfolio.

Right? Yeah. If you have no experts in diabetes or oncology, it's not like you can build it from scratch. Yeah. Overnight. So I think you, when you go to a company, you're gonna definitely inherit. What's there already? 

Armon: Yeah. 

William: Then you gotta assess like what are the strengths and weaknesses and then what are areas that you could potentially go into, but not like over 20 years, you know?

Yeah. That you could do in a reasonable amount of time. Yeah. So I think in, I mean, I think the, obviously the clinically validated targets people will accept, but there are still debates on like, you know, should we be. So like in, in Alzheimer's disease, a beta is always been a very controversial topic. 

Armon: Yeah.

William: Even though there's two medicines approved now, these antibodies, some people still feel, you know, that, that the improvement is not huge. 

Armon: Yeah. 

William: Um, and so you may still have people say like, I'm not gonna get into that space. 

Armon: and like touching on the dynamics with biotechs, I mean, what do you [00:26:00] think pharma's good at when it comes to innovation on new drugs versus what do you think biotechs are good at?

William: Yeah, that's a really good question. So now that I've been in biotech, um, I think pharma is good at actually, if their leadership is there in terms of long-term invest investment. 

Armon: Mm-hmm. 

William: So really being patient about some breakthrough molecule development. Yeah. And the examples of that would be like emicizumab.

Armon: Yeah. 

William: Which is a bispecific. Molecule antibody for hemophilia A. 

Armon: Yeah. 

William: Where one arm binds to factor nine, another one binds to factor 10, and it bypasses the need for factor vii. Yeah. I don't wanna get into too much 

Armon: detail. Yeah, yeah, yeah. 

William: But you know, two guys screened like 44,000. The scientists, two guys screened 44,000 bispecific constructs.

Armon: Yeah. 

William: Um, and you know, it was like over a period of probably 10 years before that molecule came into development. Would probably, at least today, 

Armon: they would not do that. 

William: No question. Much shorter horizon. 

Armon: Yeah. 

William: Yeah. Where they want [00:27:00] their return. 

Armon: Yeah. 

William: And so that's unlike, another example is, uh, actually from Roche was the brain shuttle.

Armon: Mm-hmm. 

William: Uh, which is actually gonna phase three, it's beta we talking about, but it took like 10 years of development to make a brain shuttle. Okay. Now there are some biotechs that are doing that now too. But you know, the initial. Investment now, now that it's been shown that the brain shuttle potentially is more effective than a lot of people are getting into there.

Armon: Yeah. 

William: So I think some of those long-term projects I think are best incubated 

Armon: in 

William: pharma. In pharma. 

Armon: Yeah. 

William: Um, and then biotech, um, you know, I think, uh, it really can fund a lot of cool innovation. 

Armon: Yeah. 

William: It seems that often very close to the academic, academic universities. Not that big. Pharma is not. The VCs have really gotten into the 

Armon: Yep.

William: Space and licensing out and trying new, promising, uh, technologies and Yeah, and therapeutic modalities and so on and so forth. So I think that's really good. Mm-hmm. I, I do. [00:28:00] Right now, we're in a period I think we're, uh. A lot of the VCs prefer closer to near term readout. 

Armon: Yeah, yeah. 

William: So want molecules closer to clinic and so on.

Yeah. But we do need to have investment in the earlier space as well. 

Armon: And and do you think like the cost of getting from, let's say, whatever, an idea for a molecule to clinical first clinical proof of concept is lower and the timelines faster for the biotechs relative to the pharmas? Or do you think that's not necessarily true?

William: Well, again, uh, let's see. Lemme think about that one. Well, I think because biotechs tend to be smaller 

Armon: Yeah. 

William: And fewer have fewer people, there's just a lot less, um, uh, networking that's needed. Yeah. And decisions can be made and 

Armon: they have a constant fear of death. So that definitely can Yeah, 

William: yeah, yeah.

Yeah, yeah, yeah. For sure. You know, existence could be gone. If the molecule doesn't make it, um, this 

Armon: tiger chasing you, you're gonna keep moving. 

William: Yeah. Yeah, yeah, yeah. Whereas in pharma, obviously if your project is dead, then you get assigned to another project. 

Armon: Yeah. Yeah. 

William: Yeah. Um, [00:29:00] yeah, so I think, again, I think in pharma, if you have the vision and the leadership that will support long-term 

Armon: Yeah.

William: You know, projects, then that's great. And then they don't have to rush as much. They have the time to really make, you know, these breakthroughs. 

Armon: Yeah. 

William: Although even with emicizumab coming back to that molecule, at one point the leadership said like. You have a year to show us that you can keep going. Yeah.

Armon: Yeah. 

William: You know, and so you just still need to have deadlines in pharma as well. Yeah. 

Armon: And, and what do you think leads to pharma doing, you know. 

What feels like relatively few biotech deals, but when they do do it, they appear to be paying such a significant premium relative to what the financial investors would've done.

William: Yeah. 

Armon: Uh, so like what do you think that dynamic is? 

William: Well, I think there's several things going on, right? So in pharma you have your own portfolio already. 

Armon: Yep. 

William: So sometimes you'll have what's called the non invented here syndrome. 

Armon: Mm-hmm. 

William: And that's really important to have your group culturally say that like, innovation can come [00:30:00] from anywhere.

Armon: Yeah. 

William: Whether it's internal or external. 

Armon: Yeah. 

William: Um, sometimes you will have to make trade-offs. 

Armon: Mm-hmm. 

William: So if you go for an external molecule, you know, then they could impact internal people. So you have to, it, it's just a consideration, that's all. Yeah. Yeah. Um, then the other thing is it's always easy to say, yeah, your idea's really cool.

I think it's promising, but show me more data. 

Armon: Yeah, yeah. 

William: Yeah. And you know, I'm, no 

Armon: one's ever said that before 

William: ec Oh, no, no. And uh, and, and, and you're willing to pay more at that point. 

Armon: Yeah. Yeah. 

William: You know, I think it's just the uncertainty. 

Armon: Yeah. 

William: Right. Um, but you would think that people, that farmers would want to invest really early.

Yeah. Uh, so that they could get it cheaper, but not, that doesn't always happen. I think they're more willing to invest in two. In my experience, you know, in a platform that looks really promising. 

Armon: Yeah. 

William: That may potentially open up a lot of avenues. 

Armon: Yeah. 

William: Um, or you know, the pharma really knows exactly what target they want to go [00:31:00] after.

Yeah. And they may even be working on internally, but then there's something externally, 

Armon: there's one multiple shots on goal. 

William: Yeah. 

Armon: Yeah. And what about the financial disparity? Because what I mean is, for example, a. Particular biotech's. Last valuation may have been a hundred million dollars, but the pharma will come in and buy it for 500 million to a billion.

William: Yeah. 

Armon: Whereas if on that exact same data, if that biotech were to raise again, there's no way in hell that would've been their valuation. 

William: Yeah. 

Armon: So why do you think either it's like. You could view it as the VCs are getting a great deal, or the pharmas are overpaying. Like who, who knows what the true price should have been?

Uh, but what do you, what do you think causes that disparity? 

William: Yeah, that's a really interesting question, Armand. So I guess every asset, uh, every asset's value depends on who wants it. 

Armon: Yeah. 

William: And so from a biotech side, I guess if you can create the demand 

Armon: Yeah. 

William: And the fomo, the fear of missing out. Yeah. Then [00:32:00] you can get a higher evaluation.

Yeah. Um, that's probably, I'm not saying it's all that. 

Armon: Yeah, yeah, yeah. 

William: But that's, uh, significantly related to that. And then, uh, as you know, also in, um, in DCS and pharma, you look at comps. 

Armon: Yep. 

William: So usually then when you have a deal. Uh, the first thing people look at is, well, what were the prior deals in that space?

That would be analogous. And then you sort of start there. So that gives you the floor and ceiling almost. Yeah. From the get go. 

Armon: Well, well, I'd also maybe argue that there's an interesting dynamic because the VCs are kind of on both sides of the table, usually because they do have their fingers already.

In that company. So they have like a mixed incentive of like, I don't wanna pay too much, I wanna leave. 'cause they're all aiming for an eventual pharma buyout. 

William: Yeah. 

Armon: Or an IPO. But you know, they kind of want that eventual pharma buyout. So they're like careful about managing price to a zone where their next delta remains huge.

William: Yes. 

Armon: Whereas often when you guys are coming and you don't have your fingers. You as a farmer, you have no fingers in there. So it's like a true opposite sides of the table [00:33:00] back and forth. 

William: Yeah. Yeah. 

Armon: So, I don't know. I feel like some of that dynamic, I, I don't know what to do about it. I'm not saying I have a solution, but 

William: yeah, 

Armon: I, I feel like that can also lead to that.

'cause I, yeah, I feel like I would always see these price disparities. I'm like, I know their valuation was that. Why did the pharma Yeah. 

William: I don't dunno if the pharma knows that, to be honest. Uh, 

Armon: I don't think they realize. Yeah, 

William: yeah, yeah, yeah, 

Armon: yeah. Because I remember we were chatting about something and you were gonna have a meeting.

I'm like, do not tell them who you are. 

William: Yeah. Yeah. 

Armon: You gotta walk in as William off the street who likes tennis. That's 

William: true. 

Armon: That's 

William: true. 

Armon: right. 

And, and so now you've actually gone to our side of the house with the biotechs. What's it been like switching over what surprised you from waltzing over to our side of things?

William: Yeah, I mean I, uh, obviously in biotech, I'm the CEO of Lio Therapeutics, you have to spend a lot more time fundraising. 

Armon: Yep. 

William: You know, in pharma, I would get a budget and justify why I should get that budget. And I have to say I [00:34:00] was able to get a, uh, increase in usually every year. 

Armon: Yeah. 

William: Unless it was like a big reorg or something like that.

Um, whereas in, uh, biotech, yeah. I'm, you know, uh, only as good as my next tranche. 

Armon: Yeah. 

William: And then have to convince the funders that I should get the next tranche or, you know, to the next raise and so on and so forth. Yeah. So a lot of time is spent on that. Um, what's interesting, I think is that, well, the other thing is I.

Roll up my sleeves all the time. 

Armon: Yeah. 

William: Uh, and so I'm doing whatever it takes, right? Yeah. I'm making slides, uh, interpreting the data, having meetings with VCs, with strategics, and so on and so forth. Yeah. So just whatever it takes because I'm invested in the company. 

Armon: Are there things you miss about the pharma side?

William: Uh, well, yes. So let me come back. So I would say, uh, I used to get a lot of, uh, complaints, I would say, from teams about the amount of slides that they had to make in pharma. 

Armon: Yeah. 

William: But I would say I would have no mercy today [00:35:00] given me a number of slides that I've had to make for the various meetings that I have.

Uh, and it's been really interesting. I think the other thing that you see is like when you meet with a lot of these various investors, you get a lot of questions. 

Armon: Yep. 

William: Whereas internal teams inside pharma, you don't really go outside of your group 'cause you don't wanna disclose anything. Yeah. So you're only meeting teams internally.

Armon: Mm-hmm. 

William: Well, most of the time. So, yeah, I'm just wondering like if I were ever to go back to pharma, like how would you replicate that kind of feed? Real time feedback, because a lot of times the feedback is really good. 

Armon: Yeah. Yeah. 

William: Like, and it got, it keeps you on your toes and stuff like that. Now, inside pharma, you may have to go to the different functions meetings.

Armon: Yeah. 

William: But it's not like the same, I think kind of experience. 

Armon: Well, well, for and for what it's worth, I don't know. I don't think it's that practical, but I do think the public side investors can be quite good. 

William: Mm-hmm. 

Armon: But I think it's just they don't normally want the r and d people to talk to them. 

William: Mm-hmm.

Armon: Uh, 'cause they're worried about, [00:36:00] are they gonna be filtered enough to talk to a public markets analyst? Because I think that feedback mechanism does exist for like a public biotech still. 

William: That's true. Yeah. 

Armon: Uh, which is like, you know, almost all the pharmas are public companies, so theoretically they can actually go do that too.

I just feel like access is way more restricted by the pharma. The analyst would love to talk to the head of r and d. Uh, 

William: yeah. Yeah. Yeah. But I don't think my internal team said Roche and Pfizer were going externally and 

Armon: Right. They're normally not allowed 

William: to. This is our program. What do you think about it?

Armon: Yeah. Normally they're not allowed to. Yeah. 

William: Yeah, yeah. 

Armon: Uh, because I think that feedback loop could exist. But that's an interesting point there. Like there's a high risk of group think within a pharma. 

William: Yeah, exactly. 

Armon: Because there's always the outside 

William: feedback, right? So I think it's important for people to think about how to get that.

Uh, 

Armon: that's an interesting point. 

William: Feedback. Um, let's see what you asked me if there's anything I miss. 

Armon: Yeah, yeah. 

William: Um, 

Armon: well, I guess aside from making less slides. 

William: Okay. No, no. I mean, I, uh, going back to what I said before, right? Uh, inside pharma is an expert for [00:37:00] everything. 

Armon: Yeah. 

William: So if you have a question 

Armon: mm-hmm.

William: You know, you don't have to go find, oh, who am I gonna talk to? Which consultant am I figure out how you just go to the person and be like, how do we deal with this issue? Right. 

Armon: Yeah. 

William: Whereas now, you know, I'm fortunate, I know a lot of people. 

Armon: Yeah. 

William: But I still have to like get a consultant potentially. Um.

Armon: Is there, is there something that's much easier on the biotech side versus the pharma side? 

William: Yeah, I think uh, the speed of decision 

Armon: making. 

William: Yeah, yeah, 

Armon: for sure. 20 that 

William: they used to deal with something this minute and the next minute it gets done. Whereas in big Pharma it may take like another, a month or something.

Armon: Yeah. Or, or if it ever happens. 

William: Yeah, exactly. 

Armon: and now that you've been experiencing it firsthand, do you see the. VCs any differently, uh, on the biotech side? Because I'm sure you had some interactions while you were in pharma. Like what seems different there? 

William: Yeah, I mean, I think I had less direct interaction with the VCs, right.

Because [00:38:00] in pharma we had BD teams. 

Armon: Yeah. 

William: And usually they would deal more with that. 

Armon: Mm-hmm. 

William: So I didn't have to consider as much, uh, yeah. You know, um, and, and yeah, I mean, we were more interested in interacting with the. Companies. 

Armon: Yeah, 

William: of course we would get updates from the various VC on what's going on in there.

Armon: Right. 

William: But still meeting with the company. 

Armon: Yeah. 

William: Yeah. So, 

Armon: well, and partially I'm wondering it in the sense of like, I feel like there's very often this pattern of certain pharma types will jump into biotech and C-level roles. Oftentimes. That'd argue having no idea what they just got themselves into.

Mm-hmm. Yeah. So if you were having a talk with one of those folks, like what is the part that you think is gonna shock them the most? 

William: Wait, say that again? Going 

Armon: through, so there's often like, I don't know, pharma BD or r and d people Yeah. That get recruited into some C-level position in biotech and I feel like many of them often have no idea what they just got themselves into.

Right. I'm not saying they're not happy in the end or sad in the end. I'm just saying like it's often a [00:39:00] shock. Uh, what are the things that you think tends to shock them most? 

William: Well, I mean, I think if it's an early stage company where the science is so critical. 

Armon: Yeah. 

William: I think the question is, you know, would they be able to direct and manage the science 

Armon: Yeah.

William: In an efficient way. Oh, 

Armon: so they like a, let's say pharma BD person and then now they're suddenly a CEO of a biotech. They're like, oh, I need to like run science. And I'd never done that. 

William: Yeah, yeah, yeah. 

Armon: Or it's been years since I did that. 

William: Exactly. Yeah. I mean, I guess, but, or, um. Yeah, that's an interesting question.

I have to think. Well, that would be the one way I'd answer that question, right? So if they were in BD and pharma 

Armon: Yeah. 

William: And then they were going to say like an early seed company. 

Armon: Yeah. 

William: Would they have the chops? I'm not, not disrespecting, I'm just saying I don't know what their background is, but if it's been more commercial or whatever, you know.

Armon: Yeah. 

William: Or they never run, uh, an RD program 

Armon: Yeah. 

William: Themselves, then would they. To do that, you'd obviously have to hire people that know what they're doing and stuff like that. 

Armon: Right, right. 

William: Yeah. [00:40:00] Um, I think the other thing is just the connections. Right? You need to have, I guess being in BD and pharma, they would probably already have the connections.

Armon: Yeah. Because I think the assumption on, I don't know, both the hiring party and them making the jump is that they have connections and when the end goal of biotech is very often a pharma buyout, because even if they go public, they still want to get bought. 

William: Yeah. 

Armon: Like very, very few of them end up.

Actually selling their own drugs and everything and being successful. So they also wanna be bought. So the assumption is, oh, they know how pharma's buying machine works. So they can work towards that end goal. Uh, and they'll have some connections to assist the other things. But then I, I would guess operationally they have no idea what life is like in this kind of raggedy, 

William: much less resources 

Armon: should done on your own.

William: I think one of the feedbacks I got also, um, when I first started my company was, well, are you willing to roll up your sleeves? 

Armon: Yeah. 

William: Because you've been in big pharma, you have an army of people helping you. 

Armon: Yeah. 

William: You know, are you gonna be able to schedule your own stuff, stuff like [00:41:00] that, move 

Armon: your own slide 

William: and people forgot, had a lab and you know, lab, you have to be very scrappy as well.

Armon: Yeah. 

William: Uh, in academia. So I guess that would be the same thing if a BD person. Uh, had a lot of people helping, you know, would they able to do, do stuff on and roll up their sleeves? Yeah. The other one I think is, at least in my experience, some of the BD folks haven't had a lot of management experience. Yeah.

They're more individual. 

Armon: Yeah. Yeah. 

William: Folks. And so when you go to a company, you know, could you manage? 

Armon: That's true. Yeah. There there'll be a lot of individual contributors. 

Armon: you know, having seen it from both sides, which not a lot of people get to see it from your vantage point, what do you think defines the, like biotechs that'll succeed versus fail?

Or like, what do you think are the most common characteristics that will define they're more likely to pully because most of them will die. 

William: I still think, uh, that it's the science that you win. Yeah. You know? So do you have a really good. Um, scientific rationale. Yeah. And data supporting your hypothesis.

And should it be better than something that's out there? [00:42:00] Uh, 

Armon: my one challenge though would be, 

William: okay. Can you sell it? 

Armon: Okay. Yeah. That last part, 

William: because 

Armon: I think there are cases that from a, let's say financial success metric or acquisition success metric. You actually never knew if this thing worked by the time it got bought.

William: Yeah. 

Armon: And people only found out after, in many of those cases it actually didn't work. 

William: Yeah. 

Armon: Uh, so it ended up, I'd argue, having nothing to do with the size. 

William: Yeah. 

Armon: But everything to do with the salesmanship. I, I know that's a little bit cynical, but, uh, I mean, do you think sometimes it'll happen that they just gotta be really good at pitching the science, but not necessarily good at executing it?

And that can be enough. 

William: I have to be

Armon: That's fine. 

William: Yeah. 

Armon: Yeah. 

William: I think, uh, yeah. Ultimately what I would say is I would hope that people are making molecules that would succeed. 

Armon: Yeah. 

William: Uh. Mainly because, well, two [00:43:00] things. One is, you know, it's a lot of investment. 

Armon: Yeah. 

William: The second thing is you are putting patients on trials for 

Armon: sure. 

William: And so, uh, I always have what's called I call the grandma test.

Armon: Yeah. 

William: Which is like what you put your own grandmother on this trial. 

Armon: Yeah. 

William: And if you wouldn't, then at least for me, then you really shouldn't be doing 

Armon: Totally agree 

William: developing that molecule. So I would just keep that standard. For myself at least. 

Armon: A hundred percent agree. 

William: Okay. 

Armon: Uh, 

William: cool. 

Armon: And then like, what do you think is gonna be the next wave of innovation coming in for our sector?

William: Well, obviously AI is really hot right now. 

Armon: Yeah. 

William: Um, and so, um, 

Armon: which part of the hotness do you agree with and which part of it you're like, eh, that's overrated. 

William: Well, I mean, I think there's a, we're in the hype base for sure. 

Armon: Yeah. 

William: And, um, I do think it will revolutionize the way we do a lot of things.

Armon: Mm-hmm. 

William: Will it revolutionize everything? I'm not sure. 

Armon: Yeah. 

William: But certainly it will speed up. Um, some aspects, you know, uh, we see already evidence that, [00:44:00] you know, you could, um, uh, screen fewer molecules, for example Yeah. Than you might have had to do in the past, getting information, obviously. Um, I think, uh, some of these new programs like alpha Fold mm-hmm.

You know, are really, um. Useful for novel ways of thinking. And then now, uh, there's many biotechs showing that they can, they say they can design molecules from ai 

Armon: Yeah. 

William: Without having to do as much in the lab. But I would say that, um, the serendipity part 

Armon: Yeah. 

William: And the unpredictability of what happens actually when you get into humans.

I'm not sure that we will address those with ai. And let me just give you some examples. 

Armon: Yeah. 

William: So like. Like when, um, the first eeg, FR tyrosine kinase inhibitors Yeah. Entered the clinic. This was in the late nineties, early two thousands. You know, we didn't even know about EGFR mutations and, uh, and they entered the clinic and.

Basically, even though there was a lot of data from the literature saying that EEG R was [00:45:00] over expressed in a lot of cancers, 

Armon: yeah, 

William: it should be applicable to a lot of different cancers. It really, they only really only worked in 10% of lung cancer patients. Yeah. And then after the fact, we found out there were EGFR mutations.

Armon: Yeah. 

William: Similarly, like aib, which is API three K inhibitor. First of all, that was the fourth molecule from uh, Novartis that made it. 

Armon: Yeah. 

William: Um, and the three other ones had died for various reasons, even though they had great preclinical data. You know, one died for tox, one died for poor pk. 

Armon: Yeah. 

William: And again, it's hard to predict when you get into humans now, I don't know, maybe AI can help improve that, but we'll have to see.

And then also PI three K mutations were discovered after the fact. 

Armon: Yeah, 

William: yeah. Right. When, so there's lots of examples of you bring a molecule into a human. You might think it worked one way. Yeah. It'll work one way, but like it surprises you. Yeah. Either from a tox perspective or from an efficacy perspective.

Armon: Yeah. 

William: And if AI's work BA based on prior knowledge 

Armon: Yeah. 

William: How is it gonna address 

Armon: well, well that and also 'cause I a hundred percent [00:46:00] agree, like the biggest. Cliff seems to be what happens when you put this in a person. 

William: Mm-hmm. 

Armon: And because all these models are as good as the data sets that train them. And we, I'd argue, measure so few things.

William: Yeah. 

Armon: In a phase one trial. 'cause a lot of it is like, you don't want to know about bad shit that's happening. 

William: Yeah. 

Armon: If you're not. Required to look at it? 

William: Well, once you get into humans, you have to do that. 

Armon: Well, I think it's more like there are certain things that they're like, I'm not gonna specifically measure that.

I'm just gonna see if it pops up in my normal tox profile, and then I'll deal with it then. So the data is often not sufficiently mechanistic and or available. 

William: Mm-hmm. 

Armon: Because like Roche sits on its own pile, like Pfizer sits on its own pile and so on. 

William: Yeah. 

Armon: Uh, these aren't necessarily all connected to each other, so I feel like there's so little detail.

Available when we are such insanely complicated systems. 

William: Yeah. 

Armon: And therefore, I'm like, how are you gonna even train a model? 

William: Well, that's what, for me, that's the biggest jump. Right. Can AI address the going from preclinical to clinical? Yeah. And right [00:47:00] now we just don't have good enough faithful human model systems.

Armon: Right, 

William: right. So there's 30 to 40 trillion cells in the human body. 

Armon: Mm-hmm. 

William: There's like, what, 20,000 genes and I, I don't know what the latest in protein coding transcripts are. 

Armon: Yeah. 

William: It used to be 90,000. I think it might be more. Then there's 8 billion people on the planet, so 

Armon: Right, right. 

William: There's no model systems right now in the lab.

Armon: Yeah. And all the chemistries that can happen because protein X and protein Y happen to be next to each other in this cell, but not in the other cell. 

William: Yeah. Yeah, yeah. Or if you have a small molecule that has a metabolite that does soandso. 

Armon: Exactly. 

William: Yeah. 

Armon: Yeah. I think it's so unpredictable, especially for small molecule.

Yeah. Biologics that might be a little bit more predictable, but, 

William: but other aspects of drug development, I think AI will be applicable to. 

Armon: Yeah. No, that be good to see, you know, you definitely have your cool new. 

Book out breakthrough. What made you wanna write that book, by the way? 

William: Yeah, actually, you know, I had a conversation with someone recently who said, oh, is this like the book you would've liked to have read when you took over the head of ki?

And actually that was a really good [00:48:00] way of sum it up. Yeah. It's like, I wish I had a book like this before I actually did drug development, so I would know more exactly what was the process and what to look out for and stuff like that. Yeah. So that is one reason. Couple other reasons. Um, you know, I wanted to highlight the unsung heroes of drug development.

Armon: Yeah. 

William: So we talked about academia and industry in drug development. A lot of academics, and again, they deserve it. Yeah. But they get a lot of the credit 

Armon: Yeah. 

William: For the initial discovery. And of course, you know, we wouldn't have a molecule without the initial discovery. Yeah. But. Getting from an idea to an actual therapeutic 

Armon: Yeah.

William: Is a whole nother level 

Armon: Yeah. 

William: Of work that is required, uh, that that requires a whole nother level of work. 

Armon: Yeah. 

William: And there's a lot of scientists that work on those and there's a lot of failure that goes into that. Yeah. So I did wanna highlight like the chemist, the toxicologists, the clinicians. Yeah. All the people that sort of contribute.

Armon: Well, no, that totally resonates because like. My exposure, I guess to that part was when I [00:49:00] was a grad student. Then I went to be a postdoc and I was starting to, I dunno, when you're a naive student there, you're like, oh, we do the hard part, the hard but glorious part of coming up with these targets or molecules or whatever.

William: Yeah. 

Armon: And then some lower life form can do the translation piece because we did the intellectually difficult part. But then the part that was making me really question that slash wanna jump to the other side, I was like, oh, there was this like nature paper that looked awesome and then I'll go ask my pi.

'cause I worked for some cool people. Like, I dunno, I remember asking Olivan, Andrea, who's my immunology postdoc supervisor, like, oh, what happened to that paper? He is like, I don't know, we published it. Postdoc lab. That's that. I'm like, okay. But yeah, like no translation. And I think he had a very good point.

He's essentially like, look. Our curr, I'm paraphrasing, but our currency is papers. If I do the next improvement or make it more drug-like that's not a nature paper. Yeah. Like no one cares. 

William: Yeah. 

Armon: So my next big paper is an any JM paper. 

William: Mm-hmm. 

Armon: Which is at least $10 million and a few years away, and we as academics are just not set up for that.

William: Yeah. 

Armon: [00:50:00] So, no, those projects are just gonna. Sit there unless pharma or someone else comes and gets it. 

William: Yeah. 

Armon: Uh, and we just move on to the next thing and hope that something happens with it. Uh, I'm like, yeah, okay. I can't do that to my project. 

William: Yeah. 

Armon: Yeah. I gotta jump 

William: Well in the book, uh, you know, we interviewed, uh, interviewed David Al.

Yeah. Yeah. Great quote about, um, you know, in academia you're building a bridge with Laco. Again, no disrespect. 

Armon: Yeah, yeah. 

William: But then getting a launch medicine is like building a bridge across the Hudson, right? 

Armon: Mm-hmm. 

William: It's like a whole nother scale. Yeah. Yeah. And also I think, uh, appreciating that even manufacturing has its own, for sure, scientific expertise.

We tend to maybe sometimes think it's less relevant, but actually being able to synthesize and manufacture a molecule at scale that can be given around the globe safely 

Armon: Yep. 

William: And so on and so forth, takes a lot of effort. Yeah, 

Armon: for 

William: sure. Which I think is underappreciated in, in medicine. 

Armon: Yeah. Well, 'cause at the lab scale you never deal with that.

Exactly. Make your tiny batch for your mice [00:51:00] or for like 10 patients. It's very different from what you said. 

William: Yeah. 

Armon: To make it consistent. 

And, and like one of the other things you were mentioning in the, you mentioned the book is around this whole institutional memory where like. They learn lessons along the way until years later, they kind of figure it out.

Is that part of what you were referring to as well as like the, oh, it turns out it was a genetic mutation we didn't know about? 

William: Yeah. I mean, well, in sort of institutional memory, so the mo, the book is about eight different molecules from eight different companies. I think what's unique about it is it not only starts with sort of the initial clinical description or the academic discovery.

Armon: Yeah. 

William: But then it actually shows what happened inside pharma or biotech 

Armon: Yeah. 

William: To actually make that medicine. 

Armon: Yeah. 

William: Right. And, and that's rare 'cause I got to interview, uh, many of the scientists inside these different companies. 

Armon: Mm-hmm. 

William: And the best example of institutional memory actually is the, uh, chapter of Paxlovid.

Armon: Yeah. 

William: Where basically the Pfizer team. Uh, had been working on the original SARS [00:52:00] virus in the early two thousands. Mm-hmm. And had already thought the three cl protease the target. Yeah. Um, the, and the virus was relevant. 

Armon: Yeah. 

William: And so they had a whole team working on it. And then when the virus disappeared, you know, the molecule got shelved.

Armon: Yep. 

William: But then say, what was it 10 years later? Yeah. 10, 15 years later when, no, 15, 20 years later. 

Armon: Yeah. 

William: Yeah. Uh, when, um, SARS-CoV-2 came around. There were like a few people left, even though the infectious disease group was disbanded. Yeah. There were a few people left that were like, oh, we worked on that thing, like, you know, many years ago.

And actually they revived the program. 

Armon: Yeah. 

William: And then had a jumpstart sort of on, uh, getting there as 

Armon: Yeah. That was super cool. 

William: Yeah. So I think the lesson there is, you know, um, the other thing you'll, you'll, uh, many people in pharma and biotech, uh, can work on molecules and drug development and never have a molecule succeed.

Yeah. I think the lesson is if you're working hard every day, you understand the principles of drug development and you know, you remember [00:53:00] your lessons learned Yeah. As you go along. Then someday, you know, serendipity may fall into your lap and they may be successful, but if you're not learning along the way, you won't be success.

Armon: Yeah. And in the meantime, they're gonna shake their fists at the head of RD that killed their program. 

William: Yeah, exactly. 

Armon: so if you, if you could wave a magic wand and change like one thing about how pharma RD works, what would you change? 

William: Oh, that's a tough question. Um, 

Armon: maybe you're allowed to change two things.

There's just limited magic in this wand. 

William: Well, yeah. I mean, okay. Un unlimited, uh, resource unlimited resources. How about that one? That's an easy one. 

Armon: Well, well, maybe joking aside, 

William: what do you, although I would say having limited resources also is very good. 

Armon: Yeah, yeah. What do you think would drive more resource into this?

Because, uh, I think an obvious comparison right now is like the insane amount of money going into AI stuff. Uh, it's just an absurd scale compared to what pharma is used to. On the [00:54:00] other hand, I'd argue on average people would rather, you know, not die or live a healthier life than I know scroll, scroll their apps, or have better purchasing suggestions for their next widget.

Um, so why do you think there is that delta in how much money we'll go into? We all use it, but relatively dumb things, but it's more like. Hundreds of billions going into, let's say AI type R and D is just a thing. 

William: Mm-hmm. 

Armon: But it's definitely not gonna happen in pharma, at least not for now. Uh, but what would it take for that to happen?

Because people clearly care about health problems. 

William: Yeah. I mean, right now, since 1945 in the US, the way the whole system has been working is that government has invested in basic science Yeah. And curiosity driven research. 

Armon: Yeah. 

William: Um, and so I think the NIH budget's like 50 billion. Mm-hmm. Um, and you know, I think pharma is not necessarily gonna do that.

I don't think the VCs will do that. 

Armon: Yeah. '

William: cause the payout is too far away. 

Armon: Yeah. 

William: You could [00:55:00] be working on a problem that has no therapeutic relevance. 

Armon: Mm-hmm. 

William: Or you could be working on a problem that like all of a sudden you're like, oh my God, this has amazing therapeutic relevance. Right. And there's so many examples of that.

Like even CRISPR was like being worked on the bacteria. 

Armon: Yeah, yeah, yeah. 

William: Right. And then once, uh, Jennifer down and Emmanuel Sharper figured out. The mechanism of CRISPR and how to make it applicable to mammalian cells. Like, you know, then like you're off to the races in terms of the therapy. So, um, 

Armon: hence the no research, no cure.

William: Exactly. Yeah. So I think, you know, I guess, well I'm not answer answering your question, but I think I am concerned about. Um, 

Armon: getting more, 

William: you know, are we gonna continue to have the investment in basic science research? 

Armon: Yeah. 

William: In the US at least. 

Armon: Okay. 

And because you've been guilty of being on both sides now, what would you change about the biotech ecosystem?

If you could change one thing, you don't get to say resources again.

William: Uh, yeah, I guess it would be, hopefully it'd be easier to raise [00:56:00] money. No, you just said, uh, 

Armon: what would improve the productivity side, because obviously like throwing more money with. Productivity constant, it'll obviously produce more. What are some of the 

William: inefficiency you noticed? Yeah. What I've seen that is that, uh, the, the VCs, you know, they get to see everything.

Armon: Yeah. 

William: Very early on. Mm-hmm. It's quite amazing. 

Armon: Yeah. 

William: Uh, from the, you know, seeding the company to the, to the, you know, the preclinical to the phase one data they see all earlier than anyone else, I think. Yeah. And so maybe figuring out how to make that system more efficient. Um, and, and, um, opening it up more, uh, 

Armon: so that more like pharmas and others can see it too.

William: Yeah, 

Armon: yeah. 

William: Yeah. 

Armon: That's a good one. Okay, cool. Any other final thoughts before we wrap? 

William: No. I want to thank you for the opportunity. It's been a lot of fun. 

Armon: Hey, thank you for coming. Appreciate it.

​ 

[00:57:00]