Latest / The BioHub - by Avetix / Episode 238 - Mandana Honu Fathom Therapeutics

Transcript

Episode 238 - Mandana Honu Fathom Therapeutics · The BioHub - by Avetix

• 0:06
  Welcome back to another episode of the Biohop Podcast.
• 0:09
  Today, I'm thrilled to be joined by Mandana Honu, who is the
• 0:13
  Chief Business Officer of Fathom Therapeutics.
• 0:18
  In today's episode, we're going to talk a little bit about
• 0:20
  Mandana's career, the journey that she's been on and the work
• 0:23
  that the folks are doing over at Fathom.
• 0:27
  Mandana, thank you so much for coming on.
• 0:29
  It's a pleasure. Thank you.
• 0:31
  Fantastic. I guess to start with Mandanic,
• 0:34
  like can you walk us through your career journey and I guess
• 0:36
  what initially drew you into biotech?
• 0:39
  Absolutely. My career journey is a bit odd
• 0:43
  because initially in high school, I wanted to be an
• 0:47
  actress and I was terrible. But I really loved people and
• 0:54
  understanding human emotion and human interactions.
• 0:57
  And that was really my passion. And I signed up for competitive
• 1:02
  acting and forensics. And I think I was very good at
• 1:06
  giving speeches, but not at pretending to be something I
• 1:10
  wasn't. And that is pretty critical.
• 1:13
  I mean, not that actors are being disingenuous, but they
• 1:17
  have to make you believe who they are, right?
• 1:20
  And I think I was always looking for some sort of career that
• 1:27
  ultimately connects people as the main value add to society.
• 1:34
  And I think part of it is that I grew up moving around a lot.
• 1:39
  So I went to five different elementary schools and had to
• 1:44
  relearn a new community every time, right?
• 1:48
  So it became very intrinsic for me to find connections with
• 1:52
  other people. And as a kid, I think I was
• 1:57
  hyper aware of feelings just as part of my personality and part
• 2:03
  of feelings was physical health conditions.
• 2:07
  So my grandmother, who I met when I was I guess 8 years old
• 2:16
  for the second time because I had moved from Iran to the
• 2:20
  United States and didn't get to see my extended family for a
• 2:24
  while until we went back. She had rheumatoid arthritis.
• 2:28
  And one of my memories from being a little kid is walking
• 2:34
  around her house and just being served the most glorious treats
• 2:38
  that you could ever imagine for, you know, an 8 year old.
• 2:41
  Ice cream, watermelon, all the snacks she would make incredible
• 2:46
  food. And what really stood out to me
• 2:49
  was that all of her joints on her hands were very crooked.
• 2:55
  And it blew my mind that someone who appeared to be in so much
• 3:01
  pain was putting so much love into physical activity, cooking
• 3:05
  for us, taking care of us. And she had so much joy.
• 3:09
  But I, you know, and now that I think about it as an adult,
• 3:12
  probably shouldn't have stared so much.
• 3:14
  But I would look at her hands and just marvel.
• 3:18
  And what was most marvelous was that she was an incredible
• 3:24
  seamstress and knitter. So she has made some of the most
• 3:27
  beautiful items of clothing that I have.
• 3:30
  And I remember a few years after that my mom got diagnosed with
• 3:36
  rheumatoid arthritis. And as a kid, her telling me
• 3:39
  that all I could imagine was my grandmother's joints.
• 3:43
  And I remember crying and being super sad about it.
• 3:48
  And what was incredible was that my mom told me, you know, since
• 3:55
  your grandmother was diagnosed, medicine has come so far that
• 4:01
  I'm never going to have my joints look like that.
• 4:04
  And she kept her promise. So she, you know, was gifted
• 4:10
  with the enbrels and humiras and methotrexate, and her management
• 4:17
  of the autoimmune disease is completely different.
• 4:19
  And that made me wonder, what if I could help in some way make
• 4:27
  medicine a reality? And so my career took a lot of
• 4:30
  turns from being pre Med at MIT and a biological engineer to
• 4:35
  realizing that most biological engineers at the time were
• 4:38
  former chemical engineers. So I stayed chemical
• 4:41
  engineering, and then I went down a path of professorship.
• 4:44
  My father's a professor in civil engineering, and I was very
• 4:47
  motivated to teach the next generation, which I still AM.
• 4:51
  And if you know me, you know, I love mentoring.
• 4:54
  But professorship felt like a scholarly endeavor that I was a
• 5:01
  bit too impatient for. And I wanted those with the
• 5:04
  patients to, you know, really cultivate that.
• 5:08
  So I turned to the business side really to put my heart and soul
• 5:15
  into connecting great science with the people that can bring
• 5:21
  it to fruition in the form of medicine.
• 5:24
  So that's kind of the journey from the beginning for me.
• 5:29
  Fantastic. Yeah, that's a, that's a great
• 5:32
  story. I didn't know that.
• 5:33
  It's super fascinating. And then from your career now up
• 5:37
  until the point you've, you know, joined Fathom, you've
• 5:39
  spent time at companies like Protellium, Saturn Adventures.
• 5:42
  Like, how have those experiences shapes the way you think about,
• 5:45
  you know, building biotech companies and, like, evaluating
• 5:48
  new science? It's a great question.
• 5:50
  I mean, every company you work with shapes how you think.
• 5:55
  Every person you work for, every CEO, every board teaches you
• 6:01
  something different. And because no two companies are
• 6:03
  the same, and I think truly the core of my answer is no two
• 6:11
  companies are the same, the same way no two people are the same.
• 6:14
  And so the biggest thing that I learned working for Portillion,
• 6:19
  the seven other biotech companies that I worked with as
• 6:24
  a fractional CBO with resilience and still as a strategic advisor
• 6:31
  to Sutter Hill Ventures, is that every problem ultimately will
• 6:42
  have a solution. And every biotech or every
• 6:47
  company all together is made-up of unique humans and will
• 6:54
  ultimately have its own unique solutions more broadly.
• 6:58
  But you can learn from them. And so one of the biggest things
• 7:01
  that I tell CE OS is when someone asks you who you want to
• 7:08
  be when you grow up as a company, try not to focus on
• 7:16
  companies that you admire as the answer.
• 7:22
  So everybody wants to be extremely successful and we say
• 7:27
  the names of these companies that we view as successful.
• 7:31
  But if I ask you, George, who do you want to be?
• 7:37
  You'll probably tell me traits that you want to have as a
• 7:41
  human. And I think that's a really good
• 7:47
  way to think about building a company is like, what are the
• 7:51
  traits I want this company to have?
• 7:53
  What are the modes of value, the friendships, the collaborations,
• 8:01
  the impact on the world? Because no company is going to
• 8:07
  be any other company, right? So, so that's really key to what
• 8:10
  I've learned, I think. I guess like we'll, we'll follow
• 8:13
  up on those points as well, which is probably what, what I
• 8:16
  guess drew you to Fathom. But just to just to start with,
• 8:18
  for those who you know aren't familiar with, with Fathom, can
• 8:22
  you just talk us through the company's approach and how
• 8:24
  you're combining physics and AI to really rethink their
• 8:27
  discovery? Absolutely Fathom is a really
• 8:34
  unique place in that the Co founders are folks that have
• 8:39
  built some of the most powerful tools in our field when it comes
• 8:48
  to protein behavior, motion, binding to drugs.
• 8:55
  And they have spent decades of their lives thinking about how
• 9:01
  binding from an atomistic level motion impacts what a drug
• 9:08
  actually does in a cell. And what drew me to the team is
• 9:14
  the dedication over a very long time to solving a problem and
• 9:23
  then continuing to level up. I think in a world where you
• 9:28
  open your phone and you really only have time to read headlines
• 9:33
  and information is shared through 4 second clips, it's
• 9:39
  really hard to maintain people's attention.
• 9:44
  And there's a deep appreciation I have for people whose
• 9:48
  attention has been on the same thing for a really long time,
• 9:52
  and they've continued to grow and build and challenge
• 9:56
  themselves more and more. And as someone who felt, you
• 10:04
  know, I was a cancer researcher for many years.
• 10:07
  I started in the lab at 16, and I ate and learned how to do
• 10:11
  molecular biology. And then I ended my career in
• 10:13
  the lab at 32. After my postdoc, I spent 16
• 10:19
  years doing research and really at some of the most incredible
• 10:24
  institutions. And I felt that biology is so
• 10:29
  complicated even if you try to do it perfectly.
• 10:33
  And so Fathom is one of the rare biotechs that accepts this on
• 10:40
  principle. And that really drew me to the
• 10:45
  Co founders is like, hey, you know, we're going to solve
• 10:48
  really, really hard problems. And we may only be solving a
• 10:52
  problem from 1 angle and we could spend a lot of time
• 10:56
  solving it from another angle. And we need to figure out what
• 10:59
  angle is going to make the biggest difference to a human
• 11:02
  patient and their actual life. And so that connection between
• 11:11
  does a molecule bind to a protein and does the protein
• 11:15
  then wiggle in the right way and then does that wiggling cascade
• 11:21
  down into more wiggling that we can anticipate in a way that
• 11:25
  affects the cell in a way that the person's actual disease is
• 11:29
  modified. That's incredible.
• 11:32
  And that is the core reason that these folks went from building
• 11:36
  software tools to working at companies like Silicon
• 11:41
  Therapeutics and Royvant to now building this company.
• 11:44
  So that's really the the premise is listen, you know, physics
• 11:50
  based models provide an incredible amount of insight
• 11:53
  into what's going on at an atomic level.
• 11:57
  But it would take a supercomputer the size of the
• 11:59
  universe billions of years to make a single calculation that
• 12:03
  is the frozen snapshot of GLP one bound.
• 12:07
  And can we start to speed that up?
• 12:12
  Well, yes, but you have to have those decades of experience and
• 12:17
  the synthetic data to actually start building something that
• 12:23
  will match reality at all. Yeah.
• 12:27
  And having spent some time with Fathom, now what?
• 12:30
  What are you most excited for, I guess over the course of the
• 12:33
  next few years at the company? I think the thing that excites
• 12:37
  me the most is truly being able to make medicine that no one
• 12:40
  else can invent because of the precision that we can achieve.
• 12:45
  And really what that means is there are two major things that
• 12:53
  everyone thinks about when it comes to a drug, regardless of
• 12:56
  whether they're a chemist or a biologist or a computational
• 13:00
  scientist. 1 is what do I want the target to do once I've
• 13:07
  manipulated it with a drug And two is how much does that need
• 13:14
  to happen for the patient to see a difference in their life, in
• 13:18
  their symptom, in their longevity, whatever it is,
• 13:22
  right? And we go about it two ways with
• 13:26
  our platform called Microcosmos. We call it Microcosmos because
• 13:31
  we want to fundamentally set the baseline that every targeted
• 13:39
  drug is in a world, in a cosmos and it cannot be siloed and
• 13:48
  thought of as a motionless entity on its own.
• 13:53
  And so there's two things that we have dedicated ourselves to
• 13:58
  that I'm most excited about. One of them is, can you start to
• 14:05
  unravel some of the hardest drug targets by understanding both
• 14:12
  atomistically what you need to do to them to get them to behave
• 14:18
  in the way that you want? So actually using chemistry to
• 14:21
  control biology. And the second thing is really
• 14:28
  being able to take a drug and start to predict every other way
• 14:36
  that it could impact the cosmos around it.
• 14:39
  Not just the target itself, but I have really critical liver
• 14:44
  enzymes that I don't want a drug to touch.
• 14:46
  Can I start to predict whether or not it would to what extent
• 14:49
  it would start designing away from that?
• 14:52
  We have this incredible example where our team asked ourselves
• 15:01
  in the midst of inventing drugs for a 20 year target that's been
• 15:09
  elusive to the community for a really long time.
• 15:13
  In the midst of building that for our pipeline, we asked
• 15:16
  ourselves, would we have been able to predict some of the
• 15:19
  drugs that are out there just based on our Microcosmos
• 15:23
  platform? Would we have predicted that
• 15:26
  they would become development candidates?
• 15:29
  And what's really cool is that we have an example of inhibitors
• 15:38
  among other examples where you can put what a single command
• 15:44
  into our tool, which is our internal facing proprietary tool
• 15:49
  and ask it make me a inhibitor for this specific target.
• 15:56
  And it will just on its own iterate four or five times and
• 16:04
  spit out not only the development candidate, but a few
• 16:07
  other options. So that development candidate is
• 16:10
  actually an approved drug. Now, the other ones are
• 16:13
  potential improvements. And what that shows you is not
• 16:17
  that we can skip all these steps and come up with a development
• 16:20
  candidate. It shows you that we're training
• 16:23
  a model that's so seamless and addresses so many points in the
• 16:29
  hit to lead development that you can actually start to truly take
• 16:38
  the human element out. And now what our humans do is
• 16:42
  they spend all of their time analyzing the data that comes
• 16:45
  out of it instead of trying to like a, you know, talented
• 16:49
  mechanic, build the connections between various things, which is
• 16:54
  a very time consuming thing. And you can only do that if you
• 16:58
  bring in the right talent. And we're lucky enough to have
• 17:01
  that. Yeah, awesome.
• 17:04
  It's it's cool to hear the way you talk about it as well,
• 17:06
  because you can, you can see the passion that that comes through.
• 17:09
  And like your last point leads me nicely onto what I wanted to
• 17:12
  talk about as well as around like obviously the hyper AI,
• 17:16
  right? It's been one of the, you know,
• 17:18
  biggest points in biotech over the last few years.
• 17:20
  Like from your perspective, you know, where is AI genuinely
• 17:24
  creating value in drug discovery today?
• 17:26
  And you know, where do you think it will go from that?
• 17:30
  It's a really good question and I think about it a lot.
• 17:34
  I think AI speeds up some of the rote work that we can do and it
• 17:44
  so that's more just from a process standpoint when it comes
• 17:48
  to proper drug discovery, I think it really depends on what
• 17:53
  you're trying to achieve. So a really cool company that
• 17:57
  I've worked with before did a computer vision test where they
• 18:04
  tried to see, OK, a human being can't tell the difference
• 18:08
  between the wells of cells that were treated with specific siRNA
• 18:15
  versus not. But I can train AAI model to do
• 18:20
  that. And they did it so so
• 18:22
  effectively that it was able to predict 100% of the time the
• 18:25
  ones that were treated because there were some pixel related
• 18:30
  distinct distinctions that enabled them to do that.
• 18:34
  That's very cool to me. But that is 1 sliver, right?
• 18:39
  And I think one of the things that we tend to do with AI is
• 18:43
  that we think we are going to be able to see the cosmos and truly
• 18:49
  we're able to see through a keyhole, a small sliver of the
• 18:55
  universe. So I think the most value of AI
• 19:01
  is in marrying things that large language models already have
• 19:07
  access to. So for example, electronic
• 19:09
  health records, things like that, where there's written
• 19:14
  material that's really excellent.
• 19:17
  I have a friend who's building a very cool company focused on an
• 19:23
  assistant to psychotherapists based on a lot of data from
• 19:30
  patient sessions. That's a very good use of AII
• 19:36
  think when it comes to drug discovery, what we need to
• 19:39
  recognize is that it depends on what part of drug discovery
• 19:43
  you're trying to solve. And ultimately the biggest
• 19:46
  impact will come from multiple platforms working closely
• 19:50
  together. And a couple examples of that
• 19:56
  are, you know, Fathom has our AI driven generative chemistry
• 20:04
  models. That are fed by a number of
• 20:10
  different synthetic experimental results and, you know, true
• 20:17
  experimental wet lab results. And we can start to find
• 20:24
  molecules that have very specific profiles much faster
• 20:28
  than a human being would be able to draw them.
• 20:31
  Yeah. Now that is really incredible
• 20:34
  because it opens the world of testing.
• 20:39
  But here's a challenge. Once you open that world, do you
• 20:43
  have the right biological assays at the right throughput to be
• 20:48
  able to test them? That's next.
• 20:51
  So that's what we're working on as well, is every phase of the
• 20:55
  drug discovery process needs to be scaled up and you have to use
• 21:00
  the right experimental approach. So one of my kind of hot takes
• 21:05
  on this is that some of the energy that goes into using AI
• 21:13
  for drug discovery ought to be redirected to trying to invent
• 21:20
  novel instrumentation to answer questions that are impossible to
• 21:24
  answer using cry OEM X-ray crystallography cell based
• 21:29
  assays. And I think some companies are
• 21:32
  trying their best to do things like that like organoid
• 21:36
  companies or companies that are trying to make virtual cells.
• 21:40
  And the truth is that they would benefit a lot from working more
• 21:46
  together. And so that, I think is core to
• 21:53
  success is collaboration and really humility in what you
• 22:01
  can't achieve with a specific platform.
• 22:03
  And I think that's a little underrated, yeah.
• 22:08
  It's an interesting take. It's almost like it's a it's a
• 22:10
  powerful tool that that complements, you know, human
• 22:13
  expertise and it kind of leads me onto you.
• 22:16
  You've mentioned a couple of time like the people at Fathom
• 22:19
  as well. How how important is the skill
• 22:22
  set of the people that have created Fathom in in making this
• 22:27
  success? I mean, I think it's everything.
• 22:34
  I think each person is truly passionate about their area and
• 22:46
  has been on a journey that has LED them into this convergence
• 22:53
  point that is fathom, which is what if I could dedicate all of
• 22:59
  my technical expertise to trying to make novel therapies come
• 23:04
  true? The hardest, most uphill battles
• 23:10
  are usually won by people who kind of Rage Against the tide.
• 23:20
  And I think the tide right now is very much in favor of both AI
• 23:25
  and undruggable targets. So we're very lucky in that
• 23:29
  sense. But going after targets and
• 23:34
  approaches that are really painstakingly chosen to make the
• 23:39
  biggest impact, even if they're extremely difficult and they
• 23:42
  have their own sort of challenges and and need for even
• 23:47
  more expertise is super worthwhile to us.
• 23:50
  And I think that mentality, in addition to the technical
• 23:54
  expertise, is what makes the Fathom team really special.
• 23:58
  Awesome. Yeah, No, it's great.
• 24:01
  Thank you for like talking us through all of that.
• 24:03
  Like it was really interesting learning about your career and
• 24:06
  then everything that you guys are doing at Fathom Mandela.
• 24:10
  To end, we always ask a very similar question on the podcast.
• 24:13
  Essentially it's you know, if you would sit down and have
• 24:15
  dinner with any scientists innovative your past or present,
• 24:20
  who would it be and why? Can I have two answers?
• 24:25
  Yeah, of course. I love this question.
• 24:30
  I haven't thought about it before, oddly, but I think a
• 24:36
  present person that I would like to have dinner with is a former
• 24:40
  professor of mine named Jane Richardson.
• 24:43
  She invented the protein ribbon diagram.
• 24:46
  She's a professor at Duke, and she and her husband, Dave teach
• 24:52
  structural biology. I don't know if they still do,
• 24:55
  but they certainly taught me in the first modelled protein that,
• 25:02
  you know, they sold the crystal structure for is actually built
• 25:07
  physically in their lab. And they're an incredible duo.
• 25:12
  You should kind of look up their story.
• 25:14
  It's really incredible. But she won a MacArthur Genius
• 25:17
  award and doesn't have a PhD in anything biochemistry related,
• 25:22
  which is really incredible. But I want to sit down with her
• 25:25
  and I want to ask her why she taught us in our first class how
• 25:30
  to edit Wikipedia and why the protein databank mattered so
• 25:36
  much to her. Because I think she's a true
• 25:39
  visionary that enabled the alpha fold, you know, outcome of being
• 25:47
  able to truly predict protein crystal structures.
• 25:51
  And if we didn't have this dedication to open science, it
• 25:57
  wouldn't exist. And I, I feel very passionate
• 26:00
  about that. And I think she's an incredible
• 26:02
  human. So it would be amazing to just
• 26:05
  ask her more questions about her life and her values and her
• 26:08
  thinking. The second person is actually
• 26:11
  someone I've thought about before in terms of meeting His
• 26:16
  name is Thomas Jennings, and he a long time ago in New York, was
• 26:21
  a tailor that invented dry cleaning.
• 26:23
  And I think he was the first black patent holder in the US.
• 26:29
  And I really want to know what it was like in a world where
• 26:34
  everyone is against you and you're the absolute first person
• 26:38
  to do something. First of all, dry cleaning.
• 26:42
  Very cool thing to invent. But also, what was it like to
• 26:47
  patent something and how well did he understand it?
• 26:50
  And I think about IP and biotech all the time and you know,
• 26:54
  there's ethical questions and, and pricing questions and all
• 26:59
  kinds of questions, but it's fascinating to think about IP
• 27:02
  and that's someone that popped into my head.
• 27:05
  Awesome. Yeah, Fantastic.
• 27:07
  Fantastic answers. No, I really appreciate it, man.
• 27:10
  Donna, thank you so much for coming on.
• 27:12
  And yeah, I look forward to getting the episode released.
• 27:16
  Really appreciate you. Thanks a million.