opensource

Keith Elliston, "Open Sourcing Medical Data"

Keith Elliston on the angel invest boston podcast.

After a lifetime of taking science from the lab to the market, repeat founder Keith Elliston has found the capstone to his brilliant career. His startup Axiomedix is supporting the creation of a massive bank of patient data tied to the genetics of each patient to enable medical research on an unprecedented scale. Imagine what could be learned by comparing the medical records of millions of patients with their genomic data! Keith also has the ambition of building blockchain technology to create a transparent marketplace where patients might make their medical data available to researchers in a safe and accountable way. Don’t miss this engrossing and accessible peek into the future of patient-centered medical discovery.

Click here to read the full episode transcript.

Here are some highlights from the podcast:

  • Keith Elliston, PhD Bio

  • “I think one of the things about that and about entrepreneurism is luck favors the prepared. You need to be in a position for these things to happen.”

  • Synopsis of Keith Elliston’s Career

  • The Collaboration between Merck & Rutgers Shaped Keith’s Thinking about Translating Findings from the Lab to Industry

  • Partnership between Merck & NIH Opened Keith’s Mind to Collaborations between For-Profits & Non-Profits

  • Keith’s First Company – Viachem Systems

  • Keith Elliston’s Take on MBA Programs

  • 9/11 Happened in the Middle of Viachem’s Series C

  • “I think one of the things that people don't ...don't understand is that the entrepreneurial element of our economy is the most tenuous with respect to economic distress.”

  • I2B2 tranSMART What it Does and How it Came About

  • “The key challenge that we see with open source is that crossing the chasm challenge.”

  • You Don’t Own Your Medical Record

  • Powerful Combination of People’s Genetic Data with their Health Records

  • The tranSMART Platform already Has 15,000 Cancer Patients on It

  • Among the Universities that Do Tech Transfer Well Keith Names Washington University

  • Keith Elliston Suggests that Big Pharma Copy the Process that Allowed Ford to Design the Mustang


Transcript of “Open Sourcing Medical Data”

Guest: Keith Elliston – “Open Sourcing Medical Data”

SAL DAHER: Welcome to Angel Invest Boston, conversations with Boston's most interesting angel investors and founders. I'm Sal Daher and my goal for this podcast is to learn more about building successful new companies, the best way I can think of doing this is by talking to people who have done it, people such as researcher and repeat founder, Keith Elliston. Keith, it's great you're here.

KEITH ELLISTON: Hi Sal, thanks for having me.

Keith Elliston, PhD Bio

SAL DAHER: It's tremendous, he completed his PhD at Rutgers University in molecular genetics. Keith has worked in the frontiers between various disciplines, which is really where most of the interesting work is done these days. Keith has also moved back and forth between the for profit and non-profit sectors, the latest of 12 companies he started offers researchers a convenient home for their patient data, we'll get more into this later on. Keith, as a service to our younger listeners, it's a standing tradition in this podcast to ask our enormously successful guests about that moment when they figured out what it is that they wanted to be doing in their life? Do you want to share that with us in your life?

KEITH ELLISTON: Well I can share some of that with you, Sal. I think for me that aha moment, the eureka moment for me was when I was an undergraduate. I had started out as a pre-med major and didn't really find my calling in medicine and then I transferred in ... was an environmental studies major and a friend of mine had recommended that I take a genetics course. So as a junior I took a genetics course and for me it was like nirvana, I had discovered this thing that was a combination of math, biology, and science that just really inspired me and after that I took every genetics course I could take. I started working for the head of the genetics department in college and was even teaching as I was an undergraduate, so for me it was really discovering the field of genetics and pursuing it from there, but it was being in a position to find a passion.

SAL DAHER: That's a great story. You know, not every person can be inspired by every single story, it's kind of a very specific Velcro, and you know, there's someone out there listening whose going to hear this and say oh wow, genetics, I never thought of that, or you know ... or just the mode of finding your way the way that you did. Much more valuable than you would suspect, because I've had people come back to me and tell me later on that it's valuable, so I keep doing it.

“I think one of the things about that and about entrepreneurism is luck favors the prepared. You need to be in a position for these things to happen.”

KEITH ELLISTON: I think one of the things about that and about entrepreneurism is luck favors the prepared. You need to be in a position for these things to happen.

SAL DAHER: Totally, totally, it is really true.

KEITH ELLISTON: Yeah, if you're not in the position, they don't happen.

SAL DAHER: Yeah, and if your mind is not open to it.

KEITH ELLISTON: Exactly.

Synopsis of Keith Elliston’s Career

SAL DAHER: You had just a remarkable life sort of going back and forth between fields and between economic sectors and so forth, so just give us a brief synopsis of your life and a sense of what drove the dynamics there?

KEITH ELLISTON: I can't say that brevity is one of my strong suits but I'll take a shot. When I was a young graduate student, I started at the University of Minnesota, which is where I'd grown up in the mid-west and my advisor had moved out to Rutgers University and he had assumed the leadership of ... it was called the Waksman Institute, and for me that was very formative. Selman Waksman was the principal investigator who led the effort that discovered streptomycin, and streptomycin was a very important antibiotic that was developed towards the end of World War Two that saved literally millions of lives and there was a young graduate student in Selman Waksman's lab that did that work, Albert Schatz. One of the very interesting elements of this is that streptomycin was discovered at Rutgers university and it was developed by a chemical company that was in New Jersey there called Merck, and Merck became as we know, a very well-known pharmaceutical company, but at the time were involved in refined chemicals.

The Collaboration between Merck & Rutgers Shaped Keith’s Thinking about Translating Findings from the Lab to Industry

KEITH ELLISTON: So, one of the key elements that was very interesting to me was this relationship between Merck and Rutgers and in fact when I joined ... when my advisor moved to the Waksman Institute, I was funded as a Merck pre-doctoral fellow and I started working giving seminars and things at Merck. So, before I'd even finished my PhD, I was doing consulting work for Merck and what was interesting was at the time, the lab I was working in at Rutgers was developing some of the early DNA sequencing methodologies. So as part of my consulting with Merck, I set up one of the first DNA sequencing labs ever in the pharmaceutical industry, that was back in 1986, so it goes back quite some time. That's a very formative experience for me because it really talks about the melding of academic science and commercialization. So rather than trying to pursue an academic career in which you really need to focus on doing something that no one else has ever done before.

Partnership between Merck & NIH Opened Keith’s Mind to Collaborations between For-Profits & Non-Profits

KEITH ELLISTON: I was really attracted to the problem-solving needs of ... more of an industrial scientific approach. How do we take these discoveries and turn them into something that would really benefit human health or human industry and that was really how I got involved in that effort? So, my formative experience as a graduate student was that initial relationship between Rutgers and Merck and in fact, I joined Merck right out of graduate school and started working with them to develop new capabilities and new technologies to aid in drug discovery. So, as we were saying, in terms of a brief synopsis, that was very formative for me, I spent ten years a Merck, I developed at Merck as an intrapreneur, developing new efforts. I started the first sequencing lab there, I started the molecular biology computing efforts there, we turned that into a bioinformatics department, and I started the genomics efforts. How I got involved in non-profit work is in fact at Merck I led a large public/private partnership between Merck, the NIH, Washington University, and others that did the first sequence characterization of about 90 percent of human genes and that was early in the genome project days.

KEITH ELLISTON: But that got me sort of started on this interplay between for profit and non- profit and I've really followed that through my career. I went from Merck, I started the genomics efforts at Bayer Pharmaceuticals, and then while I was in those efforts, I helped start a couple of companies as an advisor. One of which was Sequana Therapeutics.

SAL DAHER: Yes.

KEITH ELLISTON: That Kevin Kinsella had put together and while I was at Bayer, Malon Walton had called me up and he said I've got this great new company and it needs a chief scientist and I think you're the guy and I kept saying no, no, no, I don't want to do a startup, and he finally convinced me and I joined as the two scientists for Gene Logic, which is a company we built through IPO and was ultimately valued at about five billion before the genomics market declined, but that was my first real experience going from an intrapreneur to an entrepreneur. Since then I've basically followed up with a number of different efforts where I like to sort of dabble in the venture capital side, in the entrepreneurial side, and my main focus has been developing new companies that are really focused around an area that will have an immediate benefit either for the drug discovery process, for diagnostics, or simply aid in human health and patient treatment.

SAL DAHER: Let's go back a little bit and let’s get a little bit more fine-grained on the process that you went through when you started your first company?

Keith’s First Company – Viachem Systems

KEITH ELLISTON: Let me think about which one would be the first company. The first company that I started on my own was a company called Viachem Systems. When I left Gene Logic, one of the things that I was really concerned about in the biotech industry is a real dichotomy between the business side of biotech and the scientific side of biotech and I had taken a lesson from the technology industry, where we have technical CEO's. We have people like Bill Gates, Andy Grove, and Steve Jobs, that really had a deep understanding of the technical details of their business but also the business elements and that was and I think to a large extent is still missing in biotech and so my interest was defining whether it was easier for a scientist to learn business or to teach science to a business person. As a scientist I took the former approach to that, which is to go learn business, and I was speaking with a couple of the formative people in that stage of my career, Jean Francois Formala, that was ventures and Seth Harrison at ... whose now at Apple Tree Partners.

Keith Elliston’s Take on MBA Programs

KEITH ELLISTON: At the time at Oaken Investment Partners, I talked to them, I said should I go get an MBA, how should I learn, and Jean Francois said unless you're going to go to Columbia, don't bother with an MBA, go start a company. So I helped them start a few companies and then we were working on a project ...

SAL DAHER: What is the Columbia MBA a particularly good MBA for entrepreneurs or is it just chauvinism?

KEITH ELLISTON: I think ... well there are a couple of very good schools in my experience for MBA's, I think that Wharton is an excellent school, Columbia is a great school, Stanford is a great school. There's a little bit of excellence in a number of schools but one of the key things that I think is also present there is Jean Francois going to Columbia and felt that from a chauvinistic perspective that if it wasn't from Columbia, it didn't matter.

SAL DAHER: Right, right.

KEITH ELLISTON: But everyone was consistent in that rather than spending two years and going through an academic educational process, you'd learn much more from starting your own company.

SAL DAHER: Right.

KEITH ELLISTON: And in the process of working with Seth and Jean Francois, we had this one project that was basically going to be a spin out of Hewlett-Packard and Hewlett-Packard, before we had got that finalized, had decided to split into two companies and was completely off track. I went back and rewrote the business plan and actually recruited some of the early principle investigators, if you will, into that and we ended up starting that as a company, and so that was Viachem Systems.

SAL DAHER: Wow.

9/11 Happened in the Middle of Viachem’s Series C

KEITH ELLISTON: And so that was the first company that I as an entrepreneur did everything from the very start to get that going. We funded that through a couple of rounds of financing, we raised ... I think it was about 17 million dollars, we wrote ... it was about 5 million in revenues and then I experienced one of the things that entrepreneurs never really want to experience, which was I was in the middle of a private equity raise for Series C financing when September 11th happened.

SAL DAHER: Oh, horrible.

KEITH ELLISTON: And so the private equity markets basically dried up overnight, we were in the middle of financing and so I ended up turning around and having to repackage the business and sell it, which was very disappointing because we were on sort of this really nice trajectory toward an IPO.

SAL DAHER: That ... you know, I think it puts into context just the ... we think about the tragic loss of lives on that horrible day and then the economic impact that it had. I mean the economy completely stood still, and this is an interesting detail to note here, that because a very promising startup could not get financing, it basically sold for very little and the idea was not pursued optimally. It's just sad to see how the progress of humanity can be staggered by ignorance and folly, a very sad note but please ...

“I think one of the things that people don't ...don't understand is that the entrepreneurial element of our economy is the most tenuous with respect to economic distress.”

 

KEITH ELLISTON: I think one of the things that people don't ... perhaps entrepreneurs don't understand is that the entrepreneurial element of our economy is the most tenuous with respect to economic distress.

SAL DAHER: Yes.

KEITH ELLISTON: And September 11th taught me that ... I was running another company, a startup that we split out at Flagship Ventures called Gen Struct and when the 2008, 2009 recession happened, immediately you know, deals stopped closing.

SAL DAHER: Yes.

KEITH ELLISTON: And the pressure that you're under as a startup is immense and I think that that's the most fragile element of our economy. I think honestly from an investor perspective, from a government regulation perspective, I and more ... is we don't protect those businesses well enough and that's why sometimes I think that we have an economy and an infrastructure that does not necessarily support real innovation, but supports incrementalization.

SAL DAHER: You're spot on in this, you know. Let's think about emergency response, okay, if there's a flood, emergency response doctrine says they have to go to hospitals and they have to address the most vulnerable populations right away. That's one of the things that you know, that's sort of dogma in emergency response. Children, children in incubators, you know, old people who cannot fend for themselves and so forth, that's the people you plan for because the people who are sound of limb and can move around and so forth, they can do a lot to take care of themselves and yet when an economic emergency strikes, we help the elephants. We help the dinosaurs, we pour billions of dollars into the dinosaurs and the small fragile enterprises that are going to really shape the future of the economy, are just left to wither. Worse, they are subjected to the economic disaster of 2008, created a regulatory excess over the next few decades, their pants fell down because of overregulation. I mean that whole mess was government created, government regulation created that, I mean no sector was more regulated than the mortgage industry and banking.

SAL DAHER: Banking is really ... and what happened, banking became even more regulated, and these burdens are really tough for new enterprises. You know, if you're a two-person company, you know, one woman and one guy in a small startup, you don't have a compliance department to deal with the loads of bureaucracy that can be settled on you because some politician is trying to strut in front of the camera. It's quite the opposite, so the emergency response that they have on the economy is deadly for small enterprises. I think it's one of the reasons why more startups fail than are started now.

KEITH ELLISTON: In my experience, I didn't have an experience where government regulations had a particular impact on my business, but you know, particularly in the 2008, 2009, is ...

SAL DAHER: It's different in the financial sector.

KEITH ELLISTON: Oh, no, absolutely.

SAL DAHER: Yeah.

KEITH ELLISTON: But you know, in the biotech space, what we found is that we had customers that because of financial instability, they continued working with us but they took their payment terms from 30 days to 90 days.

SAL DAHER: Right.

KEITH ELLISTON: And for us that's a really big deal and then in trying to raise capital and raise debt, it became impossible for us to raise debt, and so doing things on that basis became impossible, which is ... again, when you're a startup, when you can leverage equity and debt, that's a great place to be leveraging your business, and those kinds of things went away, and then there's no programs of any kind. I mean the government stepped in to bail out banks and what not, which did not trickle down to small business.

SAL DAHER: No.

KEITH ELLISTON: But there was absolutely nothing to bail out small business.

SAL DAHER: They bail them out and then impose very, very rigid rules on who to lend to and so forth, which created a credit shortage. Is it any wonder that we've had a period of exceptionally slow growth, a period of exceptionally high unemployment, that lasted ... the recovery has been historically bad because of this overregulation? You know it's like to me it was a government created mess and then the government rescue just compounded the mess and they still, they're still doing the same follies. I mean don't get me started on this topic because it's ... I feel ... I came from the financial industry and I have very strong thoughts on this, but anyway, what is I2B2 tranSMART Foundation and why does it matter? This is one of the entities that you've been instrumental in setting up.

I2B2 tranSMART What it Does and How it Came About

KEITH ELLISTON: So that's a little bit of a longer story so let me just step back a little bit. In around 2011 I decided that I wanted to take a break from the venture entrepreneurial world and work in the non-profit space and my area of scientific specialty has been an area which is now called systems biology. It's an integration of genetics with molecular biology with computer science overall and aims to understand how biological systems work as a system, not as a collection of parts.

SAL DAHER: Yes.

KEITH ELLISTON: And so, I joined the CHDI Foundation to set up their preclinical drug discovery which is focused solely on Huntington's Disease.

SAL DAHER: So, what is a CHDI Foundation?

KEITH ELLISTON: CHDI Foundation ... CHDI used to stand for the Cure Huntington's Disease Initiative.

SAL DAHER: Oh okay.

KEITH ELLISTON: But they've backed away from the cure approach [crosstalk 00:16:22].

SAL DAHER: They rebranded, this is one of my pet peeves is the rebranding.

KEITH ELLISTON: Yes, so they rebranded the same brand ...

SAL DAHER: The CFA institute used to be the Chartered Financial Analyst Institute and now CFA stands for nothing, it's just ...

KEITH ELLISTON: CFA. It's a Huntington's disease virtual drug discovery organization, it's funded by ... I guess a hedge fund manager would be the right way to describe him, who funds it basically on his own.

SAL DAHER: Who is that?

KEITH ELLISTON: That would be Andrew Schechtel, from TGS Capital, and Andrew ...

SAL DAHER: Oh wow, the hedge fund managers are much maligned so it's always good to point out one that's doing great work.

KEITH ELLISTON: Well what's interesting is I've seen a number of people ... I don't know if you're familiar with Steve Cohen.

SAL DAHER: Sure.

KEITH ELLISTON: Who is funded Cohen's Veterans Bioscience.

SAL DAHER: Oh yeah.

KEITH ELLISTON: So I was involved with them when they were Orion Bio Networks. They were focused again on systems biology, but I see a number of people that have been stepping up and one of the things that seems prevalent in the non-profit world is non-profits are very hard, but if you have a singular or a small number of benefactors that get behind the mission, that makes a world of difference.

SAL DAHER: Yes.

KEITH ELLISTON: You know, the Gates Foundation, the Clinton Foundation, Cohen's Veterans Bio Science, CHDI, and others, so I think that my interest was to go in and set up a drug discovery organization from the ground up as a systems biology focused effort and I saw Huntington's disease as a unique opportunity in that there's a single causal mutation that causes Huntington's disease and so all the biology we observe and the disease downstream can be related directly back to that single cause.

SAL DAHER: Yes.

KEITH ELLISTON: And now if we can trace that pattern we can reconstruct that pattern of cause and effect relationships, we can now interrogate it for intervention points for drug discovery, so that was my interest there and over a couple of years I built up a program. We generated very large data sets, you're always finding, you know, what is the next bottleneck, what is the next limitation and for us at CHDI, once we had the data it was having a place to put it, to organize it, to analyze it. I got involved with a platform that was called TranSMART at the time, still is actually, that had been developed by Janssen, J&J Research, and that was a platform set up specifically to take this kind of what we call translational data. Data from research that we want to translate into therapeutics and I started on the board of that non-profit and over the course of things, determined that there was a real need to actually establish a ... there was a 501C6, a pharmaceutically driven trade organization.

SAL DAHER: Yes.

KEITH ELLISTON: And so, we decided that wasn't the right mechanism and we needed to establish a 501C3, so that we could access non-profit dollars and build it up as an open source software foundation. So I left CHDI in 2013 to start the TranSMART Foundation and working with colleagues, I built around that for four years, learning some of the challenges of open source software in the scientific space, which is in fact quite different than we see in the technology space, and the tranSMART platform itself had been developed from another open source product, which was called I2B2, which stands for Integrating Informatics from the Bench to the Bedside. Was developed with over 50 million dollars of NIH money through center grants and what was clear to us is that to really advance this cause of using patient data in the research paradigm to be able to more effectively discover drugs that we needed to bring this translational data and clinical data back together, so there's really a continuum.

KEITH ELLISTON: The reason that's interesting is that clinical data tends to be very simple, it's blood measurements and vital signs and things of that nature. It's like Swiss cheese in that there's lots of holes in the data and your translational data tends to be what we call high-dimensional. So, there are things like gene expression profiles, MRI images, and today the big driver is whole-genome sequence.

SAL DAHER: Yeah.

KEITH ELLISTON: And what we realized is we want to bring all this data together on a patient by patient basis and so after four years of doing this with the TranSMART Foundation, we worked with Harvard Medical School, which was running the I2B2 Foundation, to merge these two efforts together. As a part of that, the foundation’s producing a next generation platform, which I'm hoping they will rename, but currently is called I2B2 TranSMART, it's the combined product, but this is a platform which enables patient-centered outcomes research and it's been funded, in fact, a lot by PCORI, the Patient Centered Outcomes Research Institute, but it really brings ... I think if you think of it, what did Google Maps do for mapping data, right.

SAL DAHER: Right.

KEITH ELLISTON: You don't just simply look at a map and say now everything is there because you have a way of layering all this information from telephone numbers and addresses to businesses to satellite images, et cetera, that are all centered on a GPS location. We can see what that enables all the way down to self-driving cars, so what I2B2 TranSMART does is the same thing, it brings all these data together with the patient as the GPS coordinate.

SAL DAHER: Okay.

KEITH ELLISTON: And as we merged the foundations, what we realized is that for this non-profit open source platform to be adopted, it's an enterprise class platform, there needs to be a commercial entity that will provide enterprise level support. You know, Harvard Medical School and the non-profit sector don't have the resources to provide an enterprise support model for an adopter of the platform, and so we started Axiomedix, I won't say it's a spin out because a non-profit can't own equity, but as a parallel effort with the non-profit I2B2 TranSMART Foundation, to enable this platform to be adopted and integrated into the kinds of patient-centered outcomes research that will really drive personalized medicine and drug discovery over the next 20 years.

SAL DAHER: This is so ... really fascinating and extremely promising. How are you getting clinicians involved to provide the data that so important in this?

KEITH ELLISTON: Well the thing that I think is really important in precision medicine is that we need to think of at least three different groups.

SAL DAHER: Right.

KEITH ELLISTON: We need to think about clinicians because they're doing patient treatment, collecting data, et cetera, but more and more of the patients are becoming directly involved in their care.

SAL DAHER: Yes.

KEITH ELLISTON: And when we look at Fitbit's and Apple Watch's and other kinds of IoT is the patient is now generating a lot of the data that's very useful and interesting. The Michael J Fox Foundation has recently done some trials using Pebble Watches and what not and they're able to by watching the patient over time, diagnose when they're having these freeze events and that's something that the physician that sees you periodically when you walk in the office, but your devices and this IoT watches you all the time.

SAL DAHER: Right, right, well ...

KEITH ELLISTON: And so the patient is becoming much more engaged in this process and so you know, the patients and clinicians being engaged, that's really great, but that doesn't get us new drugs, that doesn't get us new diagnostics.

SAL DAHER: Right.

KEITH ELLISTON: And so it's really about bringing patients and clinicians and scientists together in the same environment with the same data, being enabled by the same data, that's going to get us new breakthroughs.

SAL DAHER: Okay.

KEITH ELLISTON: And we see that both from a clinical trials perspective but also from ... generate the kinds of real-world data that scientists and clinicians need to make better decisions and with Axiomedix, what we see is a way to start this engagement is that we didn't ... we didn't invent the product, right. This has been developed with several hundred million dollars of NIH investment and it's an open source platform. The key challenge to open source in our industry is that there's ... you’re on your own for support, so if your company adopting TranSMART or adopting I2B2, you have to hire two or three or four people that are expert in installing it, maintaining it and supporting it, loading your data, et cetera, and what we see is that's just not tenable. The early adopters will get there and if we think about Jeffrey Moore's Crossing the Chasm, the crossing the chasm challenge for open source is having that commercial support for an option.

SAL DAHER: I just want to make a comment here to our listeners, Jeffrey Moore, Crossing the Chasm, if you haven't read it, read it or at least listen to the Audible book.

“The key challenge that we see with open source is that crossing the chasm challenge.”

KEITH ELLISTON: The key challenge that we see with open source is that crossing the chasm challenge. The key thing that entrepreneurs often don't understand is the early adopter customers are willing to invest in that bleeding edge technology.

SAL DAHER: Yes.

KEITH ELLISTON: But that's not a real market.

SAL DAHER: No.

KEITH ELLISTON: And so to get to a real market you have to get to a mainstream adopter and the key gap, which Jeffrey Moore calls out in his book. The key gap is in bridging from the early adopters to the main stream and what that takes is having the right support mechanisms, the right infrastructure, installation, maintenance, et cetera, and that's usually an issue of scaling, not an issue of technology development. So with this open source platform the I2B2 TranSMART foundation has worked with government funding and with academics, et cetera, to build a really great platform, but if it takes, you know, three or five people on the customer side to install it, that's a million dollar a year expense for them, and so what we do at Axiomedix is by using you know, a cloud hosting platform to support infrastructure, et cetera, we can average those costs across a whole group of customers. So we can take that per customer cost down to less than even one FTE so they can get that platform with the full support they need without having to deploy a dedicated staff.

SAL DAHER: This is really tremendous, you're the third guest who has been working with open source software, the first was Jay Batson, on this podcast, and he was ... headed up Aquia.

KEITH ELLISTON: Yup.

SAL DAHER: The name of which evades me right now, eludes me right now, but it is basically like the back end of 60 percent of the websites.

KEITH ELLISTON: Yeah, that's Drupal.

SAL DAHER: Drupal, yes, and it's a challenging language to work with but there's a whole ecosystem of companies around it that know how to make it work and when it works, it works fantastically well. Another person was Keenan Wyrobek and he was involved with the Willow Street Garage Project for RUS, the robot operating system, which is what allowed robot components to talk to each other, and that has been fantastically successful and now I'm very interested to hear you talking about I2B2 and TranSMART and then that your ... you know, Axiomedix is taking that, you know, open source software, and then creating an environment where it's convenient to use so that it's kind of like the convenience of Microsoft but the effectiveness of Unix.

KEITH ELLISTON: Well I think it's a lot like the analytics world, where you can intake a basic Unix and run it but you have to know how to install it, support it, maintain it, et cetera, yourself, or you can buy Red Hat and Red Hat comes with ... you know. It's prepackaged, it's been made to fit your particular installation, et cetera. What I see for Axiomedix goes beyond that.

SAL DAHER: Okay.

KEITH ELLISTON: What we see is that this utility of patient data and for patients, clinicians, and scientists working together, is to democratize the environment, so today what's interesting, I don't know if you're familiar with a lot of the privacy issues going around.

SAL DAHER: Yes, yes.

KEITH ELLISTON: But in Europe they have the GDPR, which talks about data privacy and data ownership, and what's interesting is here in the United States we don't think about it but who do you think owns your patient data, you as a patient, who owns your data.

SAL DAHER: Well I'm supposed to own it but I don't have effective control of it because it's in the hands of various agents.

KEITH ELLISTON: It actually turns out that you don't own it.

You Don’t Own Your Medical Record

SAL DAHER: I don't own it?

KEITH ELLISTON: No, your provider, whoever is collecting the data, whoever stores it in a database, they own the data and they have responsibilities through HIPPA and other regulations as to what they are allowed to do with your data, but they own the data. So for example if you look at in our industry, you look at a company like 23 and Me or Ancestry.com. You send them your DNA and they sequence that for you, who owns the data.

SAL DAHER: They own it.

KEITH ELLISTON: They own the data, right, now you can sign a waiver that says you can use it for research purposes or you can sell it, which they'd like to make the default, but they have restrictions on what they can do with the data, but they own the data.

SAL DAHER: Right.

KEITH ELLISTON: And so you know, the business model for 23 and Me is not to sell lots and lots of genetic tests and make a profit, no, they sell them at a loss. Then they do 80 million dollar deals with pharmaceutical companies on the use of that data for drug discovery and what we see at Axiomedix is as we develop more of this multidimensional data, your genotype isn't really interesting without your phenotype, and that's not really interesting without you know, other kinds of data that you're going to collect.

SAL DAHER: A lot of the people in my audience are software people, explain genotype and phenotype?

Powerful Combination of People’s Genetic Data with their Health Records

KEITH ELLISTON: Well the genotype is when we think about our DNA, you know, our genetics, is the sequence of letters, A, G, C, and T, forms your genotype. So, when we do whole genome sequencing, we can look at every location in the genome and we can determine what's unique to you, what's common with the others, and that set of genes now tells us a lot about your make up. When we look at these in a population basis, what we see is common genotypes like blue eyes or blonde hair or you know, what not, and we see rare phenotypes and you know, diseases like Huntington's disease is a rare mutation, a rare genotype that we can observe through that DNA. A phenotype is now the expression of these genes, so if one looks at the gene for blue eyes, the phenotype is the actual blue eyes, the genotype is the instructions, and when we have these complex genotypes, which we all have.

KEITH ELLISTON: We have three billion base pairs in our genomes, is we want to relate this genotype to a particular phenotype, and that becomes interesting in a clinical context because all of your clinical data, whether you have high blood pressure or high blood glucose or you know, anemia, these are all phenotypes, these are things that we can measure about you. If we can relate a phenotype back to the genotype, we can now establish a relationship between those where the gene predicts the phenotype.

SAL DAHER: Right.

KEITH ELLISTON: And that's where we can now focus drug discovery efforts, et cetera, so if we look at cystic fibrosis, where we have a particular genotype, which is the ... the Delta  F508, now that's a genotype that relates to the disease and we can work on developing a drug that can moderate that disease, which is what the cystic fibrosis foundation did in working with Vertex. They developed a very genotype specific drug that treats patients with that particular condition and what we want to do is make that a universal process.

SAL DAHER: Once again I'm just very impressed with how promising all of this is, that you're taking this very rich data, which is ... I think 23 and Me is trying to do this too. I mean they're trying to get the phenotype of the ...

KEITH ELLISTON: Well what 23 and Me has is they get your genotype from your DNA.

SAL DAHER: Right.

KEITH ELLISTON: And then your phenotype, they have a web page and they'll record things that you give them, that's called patient reported outcomes, and so that's interesting in some level but patients don't really know how to diagnose themselves. So what we think is much more interesting is not the patient reported outcomes but the clinical outcomes.

SAL DAHER: Right.

KEITH ELLISTON: So what I2B2 TranSMART does is takes not just a one million snip chip, which is I believe the version 23 and Me uses today, but it takes the entire genome sequence and then it takes the expression, it takes imaging, et cetera, and all of your clinical data and it puts those together in a way that we can look at them all. So we're not relying simply on a small number of potential genotypes related to patient reported outcomes, because you're really looking at the collection of all of your clinical data and here's one of the challenges with all of your clinical data, is if I asked you for all of your clinical data, how many places would you have to go to to collect ...

SAL DAHER: Yeah.

KEITH ELLISTON: And how many of those places have all of that in one place, the answer is zero.

SAL DAHER: Yeah.

KEITH ELLISTON: And so we have this fragmentation of your clinical data, so again, one of the things that we have as a mission, is by having a common cloud based platform where we can integrate the patients clinical data with their genomotypic data, with their IoT data, is we can have a very rich collection of data on these patients. Then the issue is well, who owns the data, well different data owners own data, but we really want to empower the patient, and so what we see at Axiomedix is our stage one approach is to help hospitals and medical centers and biotech companies to put these patient data up in the I2B2 TranSMART platform, on the cloud. Then we can build applications and infrastructure to integrate across those systems to produce a single patient record, to build ... you know, in essence a federation capability. If I'm looking for a collection of patients that have, you know, Huntington's disease and high blood pressure, they are going to be in lots of different databases.

SAL DAHER: Right.

KEITH ELLISTON: So I have to be able to query across these to find those patients and then aggregate the data so that I can do analysis and so what our goal at Axiomedix is, is to initially enable people to get their data up on a cost-effective basis, but the real value of that is this next level where we can make those data centers synergistic across these platforms and we can build different kinds of products and services on top of that to enable patients, to enable clinicians, to enable scientists to very effectively use those data.

SAL DAHER: Okay, so but are you addressing it just to the clinicians or are you trying to get individuals to also opt with their data?

Keith Elliston Hopes to Use Blockchain Technology and Digital Currencies to Allow Patients to Benefit from their Health Data

KEITH ELLISTON: So our goal is to using a combination of blockchain technologies and digital currencies, is to enable the individual patient to both access and store their data and then when someone uses their data, is to be able to be incentivized for the use of their data and one of the beautiful things about digital currencies is there are auditable. So if I can come in and have someone access my data through a transaction with a digital currency, that transaction is now auditable and is a public record, so I can see how many people have accessed my data, what they did with it, et cetera. Our goal is starting with the cloud based hosted platform, but to build an ecosystem where this data can be integrated and exchanged, you know, in essence a marketplace where you don't have to define what the interaction is but you can create an area where that kind of transactional analysis can take place.

SAL DAHER: Fantastic, what a remarkable thing that would be if you managed to get that to work.

KEITH ELLISTON: Well we see this as an incremental process. I think when you're an entrepreneur you need to think in incrementalism, not always in just innovation.

SAL DAHER: No.

KEITH ELLISTON: But we think that the hosting business for us is a valuable business, it's a profitable business, or it will be a profitable business, it's not yet, but one that has value and more. We think it's going to the next level, which is where do you take things, when I have one hundred million patients worth of data in the same hosted cloud platform, what can I do with that.

SAL DAHER: Right.

KEITH ELLISTON: What next increment can I take, which will make those data valuable to all of the stakeholders, the patients, the providers, the scientists, so that we can really accelerate that discovery of new drugs and diagnostics.

SAL DAHER: Where are you now in terms of patient records?

The tranSMART Platform already Has 15,000 Cancer Patients on It

KEITH ELLISTON: So now we've just launched our tranSMART platform, we have what we call our Cancer TranSMART implementation up and running, we have a number of people working with that. That's got about 15,000 patient’s worth of data.

SAL DAHER: Okay.

KEITH ELLISTON: But we're now readying for the launch of the I2B2 TranSMART, which is really where our focus is. That platform will launch with a dataset that Harvard Medical School has loaded into the platform, it comes from the CDC, the Centers for Disease Control, and it's the National Health and Environmental Examination and Survey in NHANES, and that's 44,000 patients with 5,000 clinical variables.

SAL DAHER: Wow.

KEITH ELLISTON: So it's very, very rich data, but our goal is to work with groups like the Framingham Heart Study and others, the Fox Foundation, to get their data into these platforms.

SAL DAHER: Framingham has hundreds of thousands, right?

KEITH ELLISTON: Yeah. What we hope to do is go from tens of thousands to hundreds of thousands, to millions, to tens of millions, to hundreds of millions.

SAL DAHER: Right, right.

KEITH ELLISTON: And that's really how we measure our approach here, how successful are we, by how many patients’ data can we bring up in the same platform in an environment where they can be shared, they can be federated, and they can be leveraged and used.

SAL DAHER: And what kind of funding have you got them?

KEITH ELLISTON: Right now it's all been funded by me personally.

SAL DAHER: Oh wow.

KEITH ELLISTON: One of the advantages of being an entrepreneur if you've been around the block a few times is you can do these things.

SAL DAHER: Yes.

KEITH ELLISTON: We've gone around and looked at raising some money, I think we're still working on making sure our story is tight, and we have the right stakeholders, and I think this is a model that requires some level of investment but right now in getting things started we've been doing it right out of our own pockets.

SAL DAHER: Fantastic. Keith, let me get you on tech transfer a little bit, okay. Not any particular tech transfer office but let’s just talk about ... you've been around them a few times, I work with a lot of people who are inventors in the university environment and then take the technology out to the market. We were talking earlier and you were saying some things that were very elucidating for me so would you care to talk about that topic a little bit?

KEITH ELLISTON: Sure, sure, one of the key challenges as an entrepreneur is you're often leveraging technology, particularly in biotech, that has been developed in an academic context and what you'd like to have in those kinds of opportunities is exclusive rights to a particular technology. One of the challenges in tech transfer is that often the inventor of the technology is typically an academic, and the tech transfer office don't have a really good understanding of the real value of that invention and they also don't have a good understanding of how you realize value from the invention. When we're working in the biotech industry we have the complication of a lot of large pharmaceutical companies licensing technology where they can afford to write big checks versus small companies that don't have capitalization to write checks, but then can leverage things through royalties and other kinds of approaches.

“The most successful tech transfer offices that I've worked with have someone in a leadership role with true entrepreneurial experience…”

KEITH ELLISTON: The most successful tech transfer offices that I've worked with have someone in a leadership role with true entrepreneurial experience, where they understand how you build value in that invention, the investments you have to make in the invention to turn it into something of value, the risks that are associated, and they can structure a deal that makes sense. One of the challenges is that many universities measure their tech transfer offices solely on the amount of dollars that they raised to fund research and so they'll look at a big licensing deal where they may get a couple million dollar payment as being much better than a royalty deal that can result in maybe hundreds of millions of dollars if they did it right, but doesn't give them an upfront payment. They'll look at those as give me the cash now and I much prefer that.

SAL DAHER: You mean they have some of these licensing offices have Jay Leno's mother ... remember the story Jay Leno saying his mother would talk about steady money, actually it's the other way around.

KEITH ELLISTON: Yeah.

SAL DAHER: Jay Leno said oh my mother, she said don't take the ten-million-dollar payment today, take the thousand dollar a month payment because it's steady money. Well they are the opposite.

KEITH ELLISTON: There's a lot of wisdom to that but I think, you know, that one of the things that universities have to understand is how to we enable innovation through entrepreneurism versus simply generating value dollars from a particular investment.

SAL DAHER: Today.

KEITH ELLISTON: Today.

SAL DAHER: Right.

Among the Universities that Do Tech Transfer Well Keith Names Washington University

KEITH ELLISTON: And I think the universities that do that well, you know, I had great experiences with groups like Washington University is Saint Louis and others, when the tech transfer guy walks in and says you're going to write me a check for two million dollars, I don't have two million dollars, I'm certainly not writing you a check for two million dollars. That's a challenge but when the office engages with you and says well here's the value we see in this technology and here's a way we see to realizing that value, we want to invest in you and that as well, and so we'll give you an exclusive option to do this, you have performance requirements, right. If you don't perform then we can take it back but they can grow with you and then realize the value that technology is as it grows forward. You know, so I think it's when we look at inventions, I think we have to be very careful in how we market those and how we integrate those into an entrepreneurial activity.

SAL DAHER: It's the story of the goose that laid a golden egg.

KEITH ELLISTON: Yup, well a lot of times a big company will pay two million dollars for technology just to bury it.

SAL DAHER: Yeah.

KEITH ELLISTON: And I don't think that helps innovation in any way what so ever.

SAL DAHER: No, it does not. Keith, is there any other topic that you'd like to address, that you'd like to get into, or anything that you'd like to say?

KEITH ELLISTON: One of the things that we were talking about a little bit earlier on, one of the challenges I've seen as an entrepreneur over the past 20 years is that we tend to think we have an innovation economy but my experience is, is that we don't really recognize or facilitate true innovation.

SAL DAHER: Right.

“What we have is an economy and an infrastructure that favors incrementalism…”

KEITH ELLISTON: What we have is an economy and an infrastructure that favors incrementalism and you know, I was speaking with a friend of mine about this the other day and they said no, no, no, you're absolutely wrong. We have an incredibly innovative economy and the example they gave me was Space X. Look at Space X, they're doing amazing things, and I said really ... I love Elon Musk, I think he's a fantastic entrepreneur, but Space X is still building up to reproduce the Saturn Five missions that we flew in the 60s.

SAL DAHER: They're ... yes, Falcon Heavy is a fraction of Saturn Five.

KEITH ELLISTON: Yeah, Falcon 9 Heavy is a fraction of Saturn Five.

SAL DAHER: Yes.

KEITH ELLISTON: And what we're doing is we're transitioning the technology from government to private industry, but ... yeah, that's a great example of incrementalism.

SAL DAHER: Yes.

KEITH ELLISTON: You know and hopefully we're going to get some point where we'll have ... something goes beyond what we had 40 or 50 years ago and then you say well okay, I'll give you that one, but not to pick on Elon Musk, he says what about Tesla. I said great example, you know, Tesla did not invent the first electric car.

SAL DAHER: They're taking a bunch of little double A sized cells produced by Panasonic or somebody and making a beautiful design with it.

KEITH ELLISTON: Well you go back into some of the first cars that were ever built were electric.

SAL DAHER: Yes, yes.

KEITH ELLISTON: And so the innovation here is incrementalism again. It's can we build a bigger battery.

SAL DAHER: Right.

KEITH ELLISTON: And so what we have is even in our entrepreneurial efforts, we're favoring incrementalism and we need to think about how do we really support and spawn truly innovative efforts, you know. When we look at going from tubes to solid state circuitry, amazing.

SAL DAHER: Right.

KEITH ELLISTON: You know, to semi-conductors, amazing, right, we're looking in biology now and we're going from plasmids to CRISPR, where we can actually edit genomes, you know.

SAL DAHER: Wow.

KEITH ELLISTON: These are things that are truly innovative and what we need to think about as a society is how do we stimulate these truly innovative efforts versus these incremental efforts because in my experience, if you do something that is ... I built an AI company 15 years ago. It went out of business eventually. Today you just have to say AI and you raise money, how do you really do innovative efforts and invest in innovation and not subject it to this incrementalization.

SAL DAHER: It's culture, okay, it's very hard to change the culture of innovation that exists. I mean let’s suppose that a very beneficent billionaire, deci-billionaire, decides to spend five billion dollars working on CRISPR and her effort comes up with amazing ... just amazing discoveries, but the market is just not ready to absorb them, okay. They'll sit around for a decade before those things can be used, even someone who puts five billion dollars into something like that cannot change the natural pace of the market. You mentioned AI and your AI company 15 years ago, okay, but a moment when AI was having a momentary flourishing, it had one before in the 50s and so forth and it may be the one right now is a flash in the pan too, but you know, it's really difficult to change the way that the markets work.

KEITH ELLISTON: I think the answer to that is yes and no, I'm not sure I fully agree. I've spent 15 years of my life working in an industry that has a 15-year product cycle and that costs upwards of two billion dollars per product.

SAL DAHER: Right.

KEITH ELLISTON: How do you do that in a sustainable way and what's happened is that's the pharmaceutical industry and what we've done is developed patent extension protocols so that we can extend the life of patents, we can provide exclusivity for different elements. There are lots of things we can do as a society to benefit these long kinds of development cycles. The challenge is, is when these industries become ensconced, if we look at where the pharmaceutical industry is today, one of the key challenges is they are doing things the same way they did it 20 and 30 years ago, they're just ... doing it with a bigger hammer, right. So they're pushing a brute force method harder and harder and not focusing on innovation but what we can do is we can through various kind of efforts in investment and things like patent laws, et cetera, is we can find really good ways to stimulate truly innovative efforts and give them the time to develop so that they can become these really key things, you know.

KEITH ELLISTON: I think that the last century was really the century of physics, right. We had these amazing breakthroughs in physics, we had rocketry, we had semi-conductors, transistors, all sorts of things that happened in that world. You know with what's happened at the end of the last century and the beginning of this century has been this real focus on biology and I think we're sitting in the early stages of the century of biology and what we don't want to do is have all the innovations sequestered and inhibited by an existing infrastructure that doesn't enable us to make the investments in that biology that are going to be truly revolutionary and I think that's the challenge to us. We have an entrenched industry that has been able to manipulate government regulations and infrastructure to facilitate one particular model.

SAL DAHER: Right.

KEITH ELLISTON: But I think we have to look at ways that we can really focus on that innovative effort, things like CRISPR, to bring those forward in ways that are truly going to change the way we look at things, you know. I think biological approaches like CRISPR are going to end up being much more valuable than the chemical approaches that we've taken in the past, and so we have to build an infrastructure that's versatile.

SAL DAHER: I'm thinking of Huntington's, it's a single gene, if they can edit that gene, you got it. This brings to mind the hundreds of millions if not billion of dollars that was wasted on Theranos.

KEITH ELLISTON: Yes.

SAL DAHER: And it's ... the thought has often occurred to me, what if the money that was lavished on that folly had been spent on let’s say you have a bunch of early stage investors are putting money into an enterprise, sort of like a giant side-car fund for whatever biotech investments, early stage investors are putting money into. Basically if you have four or five early stage investors investing in something, they'll just give them a little more money, you know. If they have 500,000 dollar round, some of that could ... you know, raise it to a million and see where that takes them. You would speed up the process of discovery of this innovation, you know, something like that. I mean it just ... a little bit of spray and pray instead of these enormous billion dollar bets on single drugs that don't pan out. I mean if you're ... a huge pharmaceutical company, why don't you take a little bit of your seed corn and spray it really widely and see what effects it has.

KEITH ELLISTON: Well you raised a couple of interesting issues there. One, I think with Theranos, the key challenge there is not how much money was invested, it was several hundred million, which is not huge ...

SAL DAHER: No, not huge, I know, I know.

KEITH ELLISTON: The valuation collapse was huge.

SAL DAHER: To me that's still huge but you're right, Theranos shouldn't even be mentioned because Theranos means so many different things, absolutely, yeah.

Keith Elliston Suggests that Big Pharma Copy the Process that Allowed Ford to Design the Mustang

KEITH ELLISTON: I think when it comes to the pharmaceutical industry though, the thing that I think about is again, looking at another analogy in another industry is how have we seen innovation happen in the automotive industry and you know, I like to go back to the development of the Mustang, you know. You have these design bureaus that produce a bunch of designs and you select something and you invest in that design and if that bureau produces good designs, you go and invest again, you know. So, we do see the thing like the Lee Iacocca, right, is he gets money time and time again to go and pursue those things. I think that the pharmaceutical industry today has got the NIH syndrome, not invented here, and they're doing all this stuff internally instead of spawning these small innovative entrepreneurial efforts. I remember 10 years ago or so I was collaborating with a big pharma company and they decided they were going to reorganize to become more entrepreneurial.

KEITH ELLISTON: So what they did is they basically broke their discovery groups down into small teams and then said that they're going to base all their incentives on their success and now they can go on this new paradigm and that would automatically create these entrepreneurs. I was like ... entrepreneurs are not made, they're born, you can't take somebody who joined a big company because it's safe and because they have a pension and they have all these things and somehow restructure it to make them entrepreneurial, you know. It takes entrepreneurs in an entrepreneurial organization to be entrepreneurial.

SAL DAHER: Yes.

KEITH ELLISTON: I think that's something that these big companies have not understood, but the design bureau approach that a number of auto makers have taken, builds a small entrepreneurial organization that is risk taking and if they succeed they benefit from that risk taking and if they don't they fail, they go away. That's the essence of entrepreneurism is now they can select for these efforts that are very successful and they can get that innovative spirit. I would love to see the pharmaceutical industry do this and what they've done is almost the exact opposite, is what they've done is taken the biotech industry which I think used to have this entrepreneurial effort and they've kind of forced them into this big pharma drug discovery model, it's just that you're not inside the drug company, but you're going to follow the same model and do the same things. That's how we invest in you and when you get something valuable then we're going to buy it.

KEITH ELLISTON: I don't think that, again, that does not stimulate the innovation that we need to be really successful and different.

SAL DAHER: So that argues for a little bit of spray and pray.

KEITH ELLISTON: I think it does, yeah.

SAL DAHER: Informed spray and pray.

KEITH ELLISTON: Well you know, you water the garden and see what grows.

SAL DAHER: Yeah, yeah. Any other thoughts Keith that you want to raise?

KEITH ELLISTON: Well I think I would finish just in saying that I do think that we're on the cutting edge, the cusp of a revolution in biology and medicine and the key thing that I think is going to drive this today is the ability to collect patient data both directly from the patient, from the medical systems, from laboratories and others and utilize that data and information in a lot of different ways. In fact I don't think we can predict the different ways and so what I think we have to all focus on is how do we navigate the privacy laws, how do navigate the data ownership laws, how do we put things up in an environment that enables people to work with these things without compromising privacy, without compromising patient care, and more. I do think that what's really critically important to this is that the more data we have available on patients, on people, et cetera, is the more we have to be concerned about how that impacts their lives.

SAL DAHER: Yes.

KEITH ELLISTON: And what it should do is benefit them and not harm them and I think one of the things that we're seeing with the privacy issues throughout is that privacy is a very important thing if divulging that information is going to be harmful, but we also need to enable it to share those data in ways that's going to be beneficial. So we can't throw the baby out with the bath water and I think that's going to be one of the critical issues of our time. How we navigate this particular issue.

SAL DAHER: A really difficult challenge, no doubt. Keith Elliston, as we sum up this interview, I just want to thank you very much for coming out to our studio and sitting down with us and just talking so candidly and so wisely about this enormously consequential things that you've been working on. Thanks a lot.

KEITH ELLISTON: Thank you Sal, it's been my pleasure.

SAL DAHER: I hope our listeners have found this interview to be as rewarding as I have. Listeners, if you enjoyed this podcast, please review it on iTunes, write me at Sal@AngelInvestBoston.com with critiques or suggestions, you will find transcripts of the podcast and AngelInvestBoston.com. Do sign up for future in person events, this is Angel Invest Boston, conversations with Boston’s most interesting angels and founders. I'm Sal Daher.

SAL DAHER: I'm glad you were able to join us, our engineer is Raul Rosa, our theme is composed by John McKusick, our graphic design is by Katharine Woodman-Maynard, our host is coached by Grace Daher.