CEO Dr. John SantaLucia Emphasizes the Importance of Experiment Design in the Biotechnology and Biopharma Ecosystem.

Full transcript:

Welcome to the Radian Partners video cast series. As you know, Radian Partners is really focused on innovation. This video cast series has been now going on for the past year, amidst the COVID-19 crisis, and we’ve seen and spoken to a wide variety of companies that are really at the cutting edge of innovation, especially in the software world.

Mohammad Hamid:

Today I’m honored to be joined by Dr. John SantaLucia, who is the president and CEO of DNA Software, a very, very exciting diagnostics technology company based out of Ann Arbor, Michigan. And today we’re going to really dig into what the future of diagnostics really looks like from John’s perspective in a company right here in Ann Arbor, Michigan. Welcome to the show, John.

Dr. John Santalucia:

Thank you, Mo, for inviting me.

Mohammad Hamid:

Absolutely. So let’s start off, John, by just talking a little bit about your background. You have a very interesting career path that’s that’s led you to founding DNA Software. So let’s start there.

Dr. John Santalucia:

Okay. Sure. I started out in physics, actually, in college. Took a chemistry course, freshman chemistry course, fell in love with chemistry. And my senior year of college, I took a medicinal chemistry course. That was my first introduction to biology. And I soon discovered that I was a scientist who enjoyed all the disciplines. So, I did my graduate work at the University of Rochester in biophysical chemistry. I told people I couldn’t make up my mind, so I studied biology, physics, and chemistry.

Dr. John Santalucia:

As a professor, I remember very early on, we were doing some studies that were fundamental to understanding DNA. We were making measurements on the thermodynamics of Watson-Crick base pairing. Very basic research, but I remember like it was yesterday when I told my graduate student Nick [Baray 00:02:16] that… I said, “Hey, Nick, I think we discovered something really important here.” And he said, “Yeah, I agree.” I said, “Unfortunately, I don’t think anyone’s ever going to use it.” And he’s like, “Why? This is awesome. Predictions work as well as you can measure it.” I said, “Well, no one’s going to use it because it’s too hard. It’s just too hard.” It’s table look ups and math and adding stuff up. And I said, “We need to bring this to people’s level, if we want them to use it.” And he’s like, “Well, what can we do?” I was like, “Why don’t we make a webpage?” This was in 1994 we did that.

Mohammad Hamid:

Wow.

Dr. John Santalucia:

And the internet, among browsers, Netscape was the first browser. And I remember seeing the pictures come up slowly. But anyway, I said, “I think we should make a web server.” And he was like, “Well, how are we going to do that? We don’t know how to program it.” I’m like, “Well, we’ll teach ourselves. We’re scientists. That’s what we do. We do whatever technique will help us to solve a problem.”

Dr. John Santalucia:

So anyway, that was the genesis of me learning software. So, now I do chemistry, physics, biology, and computer science. So yeah, so we started that webpage. The first year, we had 100,000 hits. And 1994, 1995 [crosstalk 00:03:27] that was a lot of hits for a scientific website. The second year, we had over a million hits. And then I started getting a lot of requests for consulting. So I had about 30 companies or so who were asking me for consulting. It was more than I could handle. And I grew up in an environment with entrepreneurship, and I thought there’s a commercial value to this expertise that we’ve developed. And that was the start of thoughts about starting a company, DNA Software.

Mohammad Hamid:

Yeah. That’s exciting, John and you’re right. When you talk about biochemistry, microbiology, it’s a domain that a lot of people… It probably just breezes right over their head, what you’re talking about. But as we get into the story of DNA Software, I think there’s some really interesting insights into why this in particular is a domain that’s well-suited for innovation with software. So, let’s talk a little bit about that, DNA Software as a company. What’s the fundamental premise behind the company?

Dr. John Santalucia:

The fundamental premise is that many of the things that biologists do on a day-to-day basis actually have a more fundamental basis in chemistry. And so we really want to simulate all aspects of biology, and we have focused in on nucleic acids in particular, DNA and RNA. And we want to have a first principles approach to solving those types of problems. There are many other people who tried more heuristic approaches that I think have been less successful than our physics-based, chemistry-based modeling approaches.

Dr. John Santalucia:

More recently, we’ve been really interested in combining the details of the science, of how molecules interact with algorithms and databases and cloud-based computing. So the three of those things together, the science, the databases and algorithms and the high performance computing, has really changed our field and our whole approach to how we solve the problems. And it’s really fair to say that most biological applications before year 2010 or so were desktop computing applications. So we were one of the early leaders in the field of biology to go to the cloud, cloud-based computing.

Mohammad Hamid:

Sure. Sure. And in the context of software, it’s interesting because a lot of biotechnology, a lot of medical diagnostics, you think of the physical devices, you think of, as you mentioned, some of the simulation software, and all of these technologies, they fit into different parts of what I think of as this R&D life cycle, the beginning, the middle, and the end. You and I have spoken quite a bit about this. The middle and the end, they’re very busy from a technology standpoint. When it comes to operationalizing experiments, when it comes to executing experiments, when it comes to optimizing experiments, there’s a ton of different solutions out there. There’s a ton of different technologies out there, ton of different consulting solutions out there as well.

Mohammad Hamid:

But interestingly enough, in this fuzzy front end, when you’re designing the experiments, there seems to be an interesting gap in the marketplace today. Can you talk a little bit, John, about how DNA Software really addresses this experiment design gap in the marketplace, and maybe share your perspective on why this has been an ignored area. We always think of the phrase, garbage in, garbage out, but it seems like this space may have looked beyond that to some extent.

Dr. John Santalucia:

I think you really hit on something there. I think design is a truly 21st century idea. To really take a field that in the past has seemed intractable, and to really turn that into something where you can have predictability. In fact, that’s the essence of engineering, is to take known facts, put them together, make something robust. The field of biology has just been aching for that platform. It’s a real missing part. If you look at, for example, one of the things we’re experts in with design of PCR, the polymerase chain reaction, which is the basis for many of the diagnostics that we talk about for COVID and influenza A and all that sort of thing.

Dr. John Santalucia:

In the past, it was really experimental, experimentally driven field, making a new diagnostic. And the reason for that was the tools didn’t work. Honestly, if you submitted a primer design program online, you’d get a result that was as accurate as if you just tried something that you just winged it and just did it just experimentally.

Dr. John Santalucia:

The ability to simulate all of the aspects of what’s going on with a primer, all of its… What interactions are forming, what competing reactions are forming, the kinetics of the process, the thermodynamics, all those things were just missing information. There was a missing science, and there was missing data. Databases didn’t exist. And so now, that’s really part of our approach is to marry the physical sciences with the database sciences, to create a very robust design.

Dr. John Santalucia:

The other thing was people’s usages in the early 2000s here were much less ambitious than they are now. So there is a major push nowadays to miniaturize, to automate design processes, and to multiplex, to detect more than one thing at a time. As soon as you go away from just doing a single set of primers, for example, for a diagnostic, that problem scales very poorly. It’s exponentially exploding problems.

Dr. John Santalucia:

We can get to that a little later, but the computer is well suited to doing millions of experiments that can never be done in the laboratory. And 21st century algorithms allow you to even address problems that are astronomical scale, as we can get to. So all of those things. So the lack of models, the lack of data, the lack of computer resources are all exactly the three things that we are bringing to bear on modern diagnostics.

Mohammad Hamid:

Those value propositions, John, of course, they seem so relevant to everything that’s been going on over the course of the past year plus now. And it occurs to me that right now from a societal standpoint may be the most relevant time for technology like DNA Software. Can you talk a little bit about the impact that COVID-19 has had on this diagnostic space and specifically the way in which DNA Software is, I suppose, participating in the world’s, let’s call it battle against COVID-19?

Dr. John Santalucia:

It is fair to say that development of molecular diagnostics, it was an underappreciated area in terms of its impact on society. I would say that the COVID crisis has really changed people’s perception of that. A, they recognize that the tools that were at hand were insufficient to meet the need. We saw a lot of stumbling early on. If you remember, there were problems with some of the assets from the CDC. It’s not the rag on the CDC, it’s just to say they were using inferior tools. I wish they would have contacted us, actually. We could have helped them.

Dr. John Santalucia:

But that problem of people thinking that they could just do it empirically really cost many lives. And DNA Software, we recognize the problem. We’ve been in business 20 years now, and we recognize the need for diagnostics for infectious diseases, human genetic diseases, diseases involving agriculture. All of those things are areas where we can make a significant impact. And I think COVID has really focused attention, those failures, those early failures, and then even later failures of diagnostic tests with false positives of 20, 30%. False negatives, even worse.

Dr. John Santalucia:

So all of those things that we have been trying to preach to people for years that sounded boring, talking about false positives, false negatives, and limited detection, well, those are the critical things that lead to a good design or a good working assay. So, we want to get those numbers above 95%, so that we’re accuracy. And that’s achievable nowadays.

Mohammad Hamid:

You mentioned a couple of different constituencies there, John. You mentioned the government, you mentioned their collaboration with private bio pharma companies in the development of drugs and vaccines, et cetera. Earlier in the conversation, we spoke a little bit about the actual instrumentation companies in the medical diagnostics space. One thing is clear. This value chain in the diagnostic space is quite involved. And there’s lots of different constituencies in this domain. But from DNA Software’s perspective as a company, who does DNA Software serve? It seems like there’s value in a variety of different places throughout the ecosystem. But who are some of the ideal partner categories for DNA Software, both from a technology standpoint and from a services standpoint?

Dr. John Santalucia:

Our main customer base, I’d say about two-thirds of our customer base nowadays, is biotechnology companies who have an ability to stand up laboratories for doing clinical testing. They have the ability to distribute and market and all the other aspects, but developing assays is a rate limiting step for them, assays that work and that change with time. COVID is not a single disease. We’re hearing about all of the different variants of COVID, the UK variant, South African variant, the Brazil variant, and all of those are need to rapidly respond. Well, if you’re doing it in the laboratory empirically, it’s going to take you several months to develop an assay. We really want to get to a place where we have rapid response capability, where we can deliver a new design within days or faster. And in which case, the validations come back and work on the first try and get it into the clinic right away. So that’s a major opportunity area, I think, is that rapid response.

Mohammad Hamid:

There’s an old molecular biology textbook that I was peering through earlier today, just in preparation for our conversation, John, you might get a kick out of this. I was just looking at some of the terminology used and what the academics in the space had to say at the time about primer design, about experiment design, and the importance of it. And interestingly enough, a lot of the language that was used, it was pretty similar to the language that’s used in the primer design space today. What’s important is coverage, specificity, sensitivity, these sorts of things. And of course, DNA Software in part helps further calibrate and improve efforts around getting all of these different components of PCR diagnostics design honed in and tuned.

Mohammad Hamid:

But it makes one wonder, is there some notion of change in the primer design space over time, you know, lots of different industry, they may start out in the stone ages, and there’s one way of doing things. And then they evolve to improving that solution. Then ultimately, there is a sort of a denouement climax event, where technology just democratizes access to even the most complex problem solving tools. So just for our audience’s perspective, that’s trying to get some insight into how primer design, the methods, the approaches, the tools have evolved over time, what has that evolution been like from your perspective? And what’s ahead for primer design, moving forward?

Dr. John Santalucia:

All right. Well, I’ll say that there’s quite a few people who are not aware that there has been a major paradigm shift in primer design. I think a lot of folks think that primary design is submitting your primers to a program like [Alago 00:17:10] and getting back some result that is a prediction of the melting temperature, maybe using a program like [Blast 00:17:18] to look at the specificity of the primer. But meanwhile, they’re neglecting all of the things that are most important in an assay design. Some of those things are okay, but they’re really not suited to the problem. So what I’m talking about here is when a primer binds to its target, that might be where you want it to go. But in fact, that primer can bind to many other locations in the genome or in contaminating places. And those alternative binding sites can give rise to consumption of the primers, which causes a false negative in the test.

Dr. John Santalucia:

And a program like Blast is meant for looking for sequence similarity, whereas we invented a new algorithm called ThermoBLAST that actually scans primers against genomes, genome databases, looking for sequence complementarity, which is not the same as sequence similarity. So it turns out that a Blast search misses about 80% of the thermodynamically stable hits, which we catch with our algorithms.

Dr. John Santalucia:

Another thing is why do you see false negatives in an assay? Well, it could be that the target DNA or RNA has folded. Well, how do you predict the folding of an RNA or DNA? Well, we have modern algorithms called dynamic programming algorithms that are able to use our databases for thermodynamics, to predict the secondary structure of the RNA or DNA. And that competing structure, if you target a DNA strand to a region that’s folded, that’s going to make the primer very fragile and not work very well, and may cause a false negative.

Dr. John Santalucia:

On the other hand, our algorithms find the thermodynamics sweet spots, the regions of the DNA or RNA that are relatively unfolded and accessible to binding a primer. And then the additional part… That’s the science part. Then you have the database part, which essentially, to our knowledge, no other software on the market is really combining the science with the databases. So what I’m talking about here is like for SARS-CoV-2, the virus that causes COVID, collecting all of the known variants of SARS-CoV-2, and making sure that our primers bind to all known variants. So we call that a consensus primer design.

Dr. John Santalucia:

Taking into account that full database is very challenging. It’s a 21st century scaled problem, and it requires the high performance computing and deep algorithms to achieve, to achieve that marriage of the databases and the science. So we’re doing that folding, we’re doing the databases and these advanced algorithms, all of it together is required.

Dr. John Santalucia:

We actually recently published a paper with government agencies. It’s called the SPADA, the Stakeholder Panel on Agent Detection Assays. So we took that, that was a group of scientists from the FDA, the CDC, Department of Defense, and DNA Software is actually the only industrial partner in that particular publication. But we took a look, retrospectively, at the traditional methods used for primer design, versus the modern 21st century. And pretty much, probably not surprising, but the approach that we do at DNA Software really encompasses 21st century approach.

Mohammad Hamid:

Sure, sure.

Dr. John Santalucia:

To design.

Mohammad Hamid:

John, it’s January of 2021. We’re in a new year, and lots of changes on the horizon in the US, for shifting in the executive cabinet. We’re going through a lot of change as a country, and a lot of corporations are going through a lot of change in the new year, as well. As you think about 2021 and beyond for DNA Software as a company, what are some of the strategic pillars of what the world, what customers, what partners can expect to see from DNA Software moving forward, in terms of new offerings, in terms of additions to the intellectual property, or in terms of new parts of the business that you seek to add to?

Dr. John Santalucia:

Well, our first 19 years in business have been mainly focused on developing our technology. And this year, we’ve really shifted toward applying the technology. And in particular, we’ve been applying it towards some of the most challenging assays that exist in the field. So we’ve been developing multiplexed assays. So one of the early ones that we’ve done, so we were looking for partners that have the capabilities of doing large clinical distribution of these assays, but we have developed an assay first for COVID, along with flu A and flu B multiplex assay. Next, we moved on to doing other respiratory panel pathogens. So we worked with a company called InSilixa in Silicon Valley. They’ve been a great partner of ours, and we’re very happy with… We made an upper respiratory panel for them that involved 27 different viruses and three different bacterial pathogens that cause infections, upper respiratory infection.

Dr. John Santalucia:

So getting assays to all work together is very challenging. So if primer design for a single target is hard, or primer design for a single target where you have many different variants of that target is hard, these are even more challenging because now we’re doing multiple viruses, multiple bacteria. So we’ve also engaged with the Department of Defense in developing assays for bio weapons. Can’t name all of them, but some of them you would know. I could say anthrax and Ebola and things like that, where there’s a dual use possibilities there. We’ve made assays for them that are quite large. We made a 35 Plex for bacteria for them. And we were able to show on the very first try for that 35 Plex.

Dr. John Santalucia:

It would be fair to say that normally, users might see a success rating around 70% for singleplex design. So we saw 100% success for all the single plexes. So all 35 of the bacterial designs we did worked on the first try with limited detection below 50 molecules. So we hit all of the benchmarks that we wanted there.

Dr. John Santalucia:

But then in multiplex, we got 30 out of 35 to work in multiplex on the first try. And that’s a real breakthrough. That’s unheard of, to have 87% of the assays work in multiplex. And that is a tribute to the mathematical models that we’ve made and the databases and algorithms that we brought to bear on these projects.

Dr. John Santalucia:

So we were able to turn around. So we got 30 out of 35 on the first try, with no optimization whatsoever. But then after just one iteration, we were able to figure out what was wrong with those five and redesign those, and got 35 out of 35 within [crosstalk 00:24:42]. And that’s really one of our goals right now, is to be able to make it so people can iterate quickly to a full solution. So, we’ve done that for multiple partners now, where we’re making panels now for urinary tract infections, gastrointestinal infections, various blood infections right here in downtown Detroit. It’s a wonderful view we have here.

Mohammad Hamid:

Yeah, we get to hear all kinds of things outside.

Dr. John Santalucia:

So, yeah, so we’re developing a series of multiplexed assays for a wide variety of pathogens. So we’re looking at vaginitis panels, we’ve been working with Assurance Scientific Labs to develop a whole series of pathogen panels, as I’ve been talking about.

Mohammad Hamid:

Very, very helpful, John, thanks for sharing your insight. John, as we wrap up the discussion today, one thing that I wanted to get your insight on as a thought leader in the medical diagnostics domain and diagnostics design in particular, is I wanted to see if you have a call to action for the industry at large. As you mentioned, a lot of people aren’t aware of some of the technological evolutions that have happened in this space over the past several years. One might even say that still, even after the recent events of COVID-19, et cetera, there may not still be as much of an appreciation for this technology in the market as there needs to be, and a variety of other different aspects that we discussed earlier. But do you have a call to action for whether you’re a biochemist, a microbiologist, maybe you’re a contract research organization that’s trying to think about experiment design in more efficient, in more innovative ways. What would some of your thoughts be around a call to action for the industry?

Dr. John Santalucia:

I think there’s a need for changing people’s paradigms about how diagnostics should be developed. So in the past, it really was in the hands of biologists to empirically just trial and error, try to figure out how to get primers that worked. And they might be able to try 10 different primers and get something that kind of worked as a singleplex. Those fragile assay paradigm, design paradigm, is something I’d really like to have people rethink how they go about that.

Dr. John Santalucia:

And also, I’d like them to rethink their notions about the quality of predictions that are available. So many people associate primer design with programs from the 1990s, and a lot has been discovered since then, and the reliabilities have improved dramatically. And also, I’d like them to be not so limited in their thoughts about just doing a singleplex. Why not design an assay that is more affordable, that addresses a larger market, and is more targeted to the reality that is present in the clinic? And I’m talking to here about detecting all the different variants, or detecting antimicrobial resistance variants, or detecting multiple viruses or bacteria in a single test.

Dr. John Santalucia:

So I would say that I would like to see people change their biases. So probably the number one problem we have when we address customers is that they’re like, “Well, this isn’t how we have done it in the past. We’d have the molecular biologists just order some primers and try it until they succeeded.” Well, oftentimes, they’d just fail outright. And now, that’s just not a rational approach. Why would you do that, when you have at hand a method that is a simulation optimization method that can truly give you outstanding designs on the first try? And particularly some of the more complicated assays.

Dr. John Santalucia:

So I would say that the thought leaders in this field, the companies that are leading the field are the ones who are releasing themselves with the shackles of the previous generation of design paradigm. And they’re instead coming to companies like DNA Software that have unique expertise and ability to deliver world-class assays on the first try.

Mohammad Hamid:

Absolutely, John. And that actually brings us to the conclusion of our talk today. If there’s one thing that’s clear to me, John, it’s that in a space where accuracy is so important, in a space where reliability is so important, that can result in life-changing results at a global scale, this is the kind of technology that the ecosystem needs to be investing in. Whether you’re on the biotech side, the bio pharma side, whether you’re on the CRO side, getting smarter about experiment design really is the future about when it comes to medical diagnostics and companies like DNA Software, which there aren’t many of, are really paving the way for the future. So, thanks again, John, for joining me today, and looking forward to talking to you again.

Dr. John Santalucia:

All right. Thank you for having me. I really appreciate it. It was fun.

Mohammad Hamid:

Thanks, John.

#Panelplex #Multiplepcr #multiplex #assaydesigns #howtodesignPCR #primerdesign #covid19 #globalpandemic #dnasoftware #singleplex #vaccines #pfizer #moderna #sarscov2 #herdimmunity #viruses #science #research #biotech #genomics #CRISPR #genesplicing #probes #primers #insilico #newstrain #coronavirus #masks #2021 #reddit #DNA #RNA