{"id":572181,"date":"2019-03-13T08:00:06","date_gmt":"2019-03-13T15:00:06","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?p=572181"},"modified":"2020-04-23T14:55:09","modified_gmt":"2020-04-23T21:55:09","slug":"programming-biology-with-dr-andrew-phillips","status":"publish","type":"post","link":"https:\/\/www.microsoft.com\/en-us\/research\/podcast\/programming-biology-with-dr-andrew-phillips\/","title":{"rendered":"Programming biology with Dr. Andrew Phillips"},"content":{"rendered":"<div id=\"attachment_572184\" style=\"width: 1034px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-572184\" class=\"wp-image-572184 size-large\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-1024x576.png\" alt=\"\" width=\"1024\" height=\"576\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-1024x576.png 1024w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-300x169.png 300w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-768x432.png 768w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-1066x600.png 1066w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-655x368.png 655w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-343x193.png 343w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788.png 1400w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><p id=\"caption-attachment-572184\" class=\"wp-caption-text\">Head of Biological Computation Group Dr. Andrew Phillips<\/p><\/div>\n<h3>Episode 67, March 13, 2019<\/h3>\n<p>When we think of information processing systems, we often think of computers, but we ourselves are made up of information processing systems \u2013 trillions of them \u2013 also known as the cells in our bodies. While these cells are robust, they\u2019re also extraordinarily complex and not altogether predictable. Wouldn\u2019t it be great, asks <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/aphillip\/\">Dr. Andrew Phillips,<\/a> head of the <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/group\/biological-computation\/\">Biological Computation Group at Microsoft Research in Cambridge<\/a>, if we could figure out exactly how these building blocks of life work and harness their power with the rigor and predictability of computer science? To answer that, he\u2019s spent a good portion of his career working to develop a system of intelligence that can, literally, program biology.<\/p>\n<p>Today, Dr. Phillips talks about the challenges and rewards inherent in reverse engineering biological systems to see how they perform information processing. He also explains <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/blog\/scientists-discover-how-bacteria-use-noise-to-survive-stress\/\">what we can learn from stressed out bacteria<\/a>, and tells us about <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/news.microsoft.com\/innovation-stories\/station-b\/\">Station B<\/a>, a new end-to-end platform his team is working on that aims to reduce the trial and error nature of lab experiments and <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/blog\/researchers-build-nanoscale-distributed-dna-computing-systems-from-artificial-protocells\/\">help scientists turn biological cells into super-factories<\/a> that could solve some of the most challenging problems in medicine, agriculture, the environment and more.<\/p>\n<h3>Related:<\/h3>\n<ul type=\"disc\">\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/podcast\">Microsoft Research Podcast<\/a>: View more podcasts on Microsoft.com<\/li>\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/itunes.apple.com\/us\/podcast\/microsoft-research-a-podcast\/id1318021537?mt=2\">iTunes<\/a>: Subscribe and listen to new podcasts each week on iTunes<\/li>\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/subscribebyemail.com\/www.blubrry.com\/feeds\/microsoftresearch.xml\">Email<\/a>: Subscribe and listen by email<\/li>\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/subscribeonandroid.com\/www.blubrry.com\/feeds\/microsoftresearch.xml\">Android<\/a>: Subscribe and listen on Android<\/li>\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/open.spotify.com\/show\/4ndjUXyL0hH1FXHgwIiTWU\">Spotify<\/a>: Listen on Spotify<\/li>\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.blubrry.com\/feeds\/microsoftresearch.xml\">RSS feed<\/a><\/li>\n<li><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/note.microsoft.com\/ww-registration-microsoft-research-newsletter-s.html?wt.mc_id=S-webpage_podcast\">Microsoft Research Newsletter<\/a>: Sign up to receive the latest news from Microsoft Research<\/li>\n<\/ul>\n<hr \/>\n<h3>Final Transcript<\/h3>\n<p>Andrew Phillips: So, what Station B is aiming to do is to develop a platform, a system, that will transform programming biology from what is currently a process of trial and error to something that\u2019s systematic and predictable. And that requires bringing together many different pieces of the puzzle. In programming biology, there\u2019s this sort of standard \u201cdesign, build, test, learn\u201d cycle. So, we\u2019re trying to combine these different stages of programming into an integrated platform.<\/p>\n<p><strong>Host: You\u2019re listening to the Microsoft Research Podcast, a show that brings you closer to the cutting-edge of technology research and the scientists behind it. I\u2019m your host, Gretchen Huizinga.<\/strong><\/p>\n<p><strong>Host: When we think of information processing systems, we often think of computers, but we ourselves are made up of information processing systems \u2013 trillions of them \u2013 also known as the cells in our bodies. While these cells are robust, they\u2019re also extraordinarily complex and not altogether predictable. Wouldn\u2019t it be great, asks Dr. Andrew Phillips, head of the Biological Computation Group at Microsoft Research in Cambridge, if we could figure out exactly how these building blocks of life work and harness their power with the rigor and predictability of computer science? To answer that, he\u2019s spent a good portion of his career working to develop a system of intelligence that can, literally, program biology.<\/strong><\/p>\n<p><strong>Today, Dr. Phillips talks about the challenges and rewards inherent in reverse engineering biological systems to see how they perform information processing. He also explains what we can learn from stressed out bacteria, and tells us about Station B, a new end-to-end platform his team is working on that aims to reduce the trial and error nature of lab experiments and help scientists turn biological cells into super-factories that could solve some of the most challenging problems in medicine, agriculture, the environment and more. That and much more on this episode of the Microsoft Research Podcast.<\/strong><\/p>\n<p><strong>Host: Andrew Phillips, welcome to the podcast.<\/strong><\/p>\n<p>Andrew Phillips: Thanks for having me. It\u2019s a pleasure to be here.<\/p>\n<p><strong>Host: So, you\u2019re the head of the biological computation group at MSR in Cambridge. What\u2019s going on in your group? Broad strokes, what big problems are you trying to solve? What gets you up in the morning?<\/strong><\/p>\n<p>Andrew Phillips: Well, one of the things we\u2019re really working on is trying to understand how biological systems compute. So biological systems, like living cells, they actually perform information processing, but they compute via means that we don\u2019t quite fully understand. So, my team is working on trying to reverse engineer how these living systems, biological systems, perform information processing.<\/p>\n<p><strong>Host: I want to unpack the big suitcase of biological computation a bit more. As you\u2019ve noted, programming biology\u2026 it\u2019s now new. But there are some new things that are going on. So, give us a short primer on biological computation. Where did this all get started and why, and where are we today?<\/strong><\/p>\n<p>Andrew Phillips: So, one of the things we\u2019re focusing on is trying to understand how to program biological systems. But to do that, you need to understand how these systems function. If you wanted to try to fix a car, but you didn\u2019t understand how the components worked, you couldn\u2019t just randomly change those components and expect the car to work. So, in order to program a system, we need to understand how it\u2019s working and, in our case, we need to understand how these cells are computing. Now, as a species, we\u2019ve been using cells to do things for us for thousands of years. We\u2019ve used yeast to make bread or to brew beer. And then, several decades ago, we were able to reprogram microorganisms to produce medicines, things like insulin. And now we\u2019re taking this to the next level as a field, as a discipline, and programming organisms to do much more sophisticated things, make much more complex medicines, fuels and materials. So, understanding how systems compute is an important step to being able to program them more effectively.<\/p>\n<p><strong>Host: When did somebody discover, hey, we could actually make this do something for us or make it do something different than it does? Was that an \u201caha\u201d moment or an accident, or did people try to start manipulating biology?<\/strong><\/p>\n<p>Andrew Phillips: Well, I think the first example of programming a microorganism to make a medicine, in this case, insulin, was in the seventies. And this was as we started to understand more about DNA, about how cells work. And then, more recently, we\u2019ve been able to sequence the human genome. And now we\u2019ve been able to write DNA. We\u2019ve been able to write genes. So, there\u2019s been this steady progress in technology that sort of underpinned our ability to program biology. And in fact, there\u2019s been an exponential growth in our ability to read DNA and also to write DNA. And then, more recently, we\u2019ve had some transformations in our ability to edit DNA through things like <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/project\/crispr\/\">CRISPR<\/a>. So, we have this underlying technology that\u2019s allowing us to manipulate DNA, read, write and edit it. And that\u2019s also underpinned this technological growth in our ability to program biology.<\/p>\n<p><strong>Host: Well, let\u2019s talk about those underpinnings for a minute. There are some pieces that need to be in place before we can make significant progress. You\u2019ve alluded to DNA. Are there other pieces that we really need to understand before we can move forward and make biology work for us more specifically? And even more predictably and less expensively?<\/strong><\/p>\n<p>Andrew Phillips: Well, yeah, I mean, there\u2019s still a lot we have to learn in terms of understanding how biological systems function. So, biological systems are highly complex, they\u2019re massively parallel, they\u2019re probabilistic. In many ways, they\u2019re closer to analog computing systems than the digital ones that we\u2019re familiar with. So, we still have a lot of work to do to reverse engineer these systems. So that\u2019s a challenge, understanding how these systems work. Another challenge is that we still lack a way of doing biological experiments systematically and reliably. A lot of experiments are done manually, they\u2019re time-consuming, they\u2019re error prone. And in fact, recent studies have shown that most biological experiments are not even reproduceable. And then the final challenge is that we actually lack the technology stack for programming biology. There isn\u2019t really a systematic way. In many ways, programming biology is sort of similar to the early days of trying to program silicon before the advent of high-level languages and the fundamental theory of computing that we sort of take for granted today. So, we\u2019re sort of still in the days of almost punch cards and very basic programming technology.<\/p>\n<p><strong>Host: That\u2019s funny. So, you alluded, just now, to our much more advanced ability to read and write DNA. How has that impacted the growth in programming biology, and what limitations do we still face, aside from the things that you\u2019ve just mentioned in terms of what we don\u2019t understand?<\/strong><\/p>\n<p>Andrew Phillips: Yeah, so this technology has been hugely important and has enabled the progress that we\u2019ve seen to date in programming biological organisms. So, by that, I mean reading, writing and editing DNA. But on its own, it\u2019s not enough. So, we can read an entire genome, but we still don\u2019t understand what most of it means. And we can write an entire gene, but we\u2019re still unable to predict how that gene will behave inside a living organism. And we can now edit DNA with really high precision, with technologies like CRISPR, but we\u2019re still unable to predict the consequences of those edits. So really, we\u2019re still in a situation where programming biology is done by trial and error.<\/p>\n<p><strong>Host: So, what is it about biological systems that confounds our ability to program them? We\u2019re coming at this from a computer science angle, so we\u2019re basically talking about using programming languages to compile biological algorithms to DNA code instead of binary. Talk about the differences between how biological cells operate and how computer programs operate. What are the unique challenges that scientists face in programming biology?<\/strong><\/p>\n<p>Andrew Phillips: So yeah, essentially, biological programs operate in a fundamentally different manner to traditional silicon-based programs that we\u2019re used to writing. So, you can think of a traditional computer program more like a recipe where you have a list of actions that happen in a particular order. You know, do step one, then step two. And maybe you\u2019ll repeat this N times. Whereas biological systems, they actually compute via fundamentally different means. So, it\u2019s more like a chemical soup where you have thousands of proteins interacting in parallel in a noisy fashion, and many of these interactions can go wrong with some probability. But yet out of all that noise emerges a fairly robust algorithm that is used to compute things like, when should a cell divide? Or how should an immune system respond to a foreign invader? Or even things like the internal body clock, which is essentially a combination of genes and protein interactions that computes a 24-hour period fairly reliably. So, these algorithms are actually very complicated for us to understand because we\u2019re not used to that. We\u2019re still trying to reverse engineer them.<\/p>\n<p><strong>Host: Let\u2019s talk about noise for a second. You\u2019ve just mentioned it, and you\u2019ve recently published a paper about how bacteria use noise to survive stress. So, tell us about this. What insights did you gain from this research about noise and bacteria? And what are the implications for the work that you\u2019re doing?<\/strong><\/p>\n<p>Andrew Phillips: So, this is one of many examples, actually, of how we, as a team at Microsoft Research, are collaborating with leading scientists in many different fields in universities. So, this particular collaboration with The University of Cambridge was with James Locke, the Department of Biochemistry Sainsbury Laboratory, and a joint PhD student, Microsoft funded, Om Patange. And we were looking together at trying to understand the role of noise in how bacteria survive stress. Now, stress, in this case, is not an emotional response. It\u2019s more about, you know, if you are the bacteria are in adverse conditions, so you give them hydrogen peroxide or some kind of dangerous compound that could potentially kill them, how do they survive? And this work, you know, Om did most of the experiments for this and we looked together at the computational modeling side, is trying to understand how bacteria can actually anticipate stress and actually survive. And it turns out that bacteria are growing in a noisy fashion, and they\u2019re also turning on a stress response sort of randomly. And this noisy growth and noisy stress response are coupling so that bacteria that are growing slowly are actually more able to survive the stress and also some fraction of the bacteria randomly decide to get into this state so that if a stress happens to be applied in the future, they actually survive. And so, this is a really kind of interesting example of how noise can perform a useful function for bacterial systems. But more generally, it\u2019s one of the examples of how we\u2019re trying to understand the mechanisms that bacteria and other living cells use in order to survive and process information more generally.<\/p>\n<p><strong>Host: So, you\u2019re talking about bacteria, writ large, and we know that some bacteria is actually really, really bad. And we don\u2019t want that to survive. Is there a way to parse out, hey, I\u2019m going to, you know, provide some noise and stress to the bacteria that I want to survive?<\/strong><\/p>\n<p>Andrew Phillips: Well, yeah, it\u2019s essentially trying to understand how a system works. Then we can try and direct it, reprogram it, depending on what we want it to achieve. So, if it\u2019s a dangerous infection that we\u2019re trying to eliminate, then we can understand where we want to perturb that system. For example, trying to overcome things like antibiotic resistance. And if it\u2019s a beneficial bacteria, for example, the bacteria that lives inside our gut, we want those bacteria to survive because they provide tremendous benefits to us. And so, understanding the mechanisms that bacteria use in general can help us determine what strategies to use in the beneficial case and in the harmful case.<\/p>\n<p>(music plays)<\/p>\n<p><strong>Host: What are the most promising applications of the research you\u2019re doing? What\u2019s is the field hoping for, before we start talking about some specific things that you are doing at Cambridge?<\/strong><\/p>\n<p>Andrew Phillips: Yeah, it\u2019s a really exciting field. It\u2019s often referred to as synthetic biology, where the goal is to program biological systems more systematically using engineering-based principles. And so, this field, as a whole, is moving forward rapidly and there are many applications that are actually currently making excellent progress, and there are many potential future applications. The ones that excite me most are actually in the medical field. Biologics, these are drugs made by reprogrammed organisms. And they essentially are the fastest-growing sector in the pharmaceutical industry, and they account for over half of industry revenues and annual drug approvals. And they\u2019re actually some of the most powerful treatments we have for diseases like cancers that many traditional drugs, chemical-based drugs, are not able to treat. And so, these biologics, they\u2019re too complex to be made by ordinary chemical means. And so instead, they\u2019re made by genetically programmed organisms that act as living factories. And biologics also includes sort of more advanced treatments. One example is cell therapy, where you can actually reprogram a patient\u2019s immune cells to target specific cancers, and there\u2019s an example of a company, Oxford BioMedica, with whom we\u2019re working, that, in partnership with Novartis, they\u2019ve developed the first living cancer drug which essentially reprograms a patient\u2019s immune cells to fight cancer with 80 percent patients in complete remission in the first trials. So that\u2019s one of the most exciting areas. But there are also many other areas. Agriculture is another one. So, nitrogen fertilizer is responsible for five percent of global greenhouse emissions, and half of the fertilizer is washed away causing toxic pollution. And this company called Pivot Bio, they\u2019ve essentially reprogrammed soil microbes to transfer nitrogen directly to the plant roots without emitting these greenhouse gases and with almost no pollution. So very little is washed away. These programmed microbes are actually performing extremely well in recent trials in the field. And then there\u2019s a lot of potential for other industries as well, like construction. So, the cement industry accounts for about five percent of global carbon dioxide emissions. And there\u2019s a company called bioMASON that\u2019s reprogrammed microbes to produce cement at ambient temperatures so they can get rid of most of these emissions. And then textiles. So, the textile industry generates about a fifth of the world\u2019s industrial water pollution mainly in developing countries, and this company called Colorifics, it\u2019s an early startup, but they\u2019ve actually programmed microbes to produce and fix dyes to fabric using ten times less water than traditional dying methods. And then, there are a whole load of other examples. For instance, in the chemical industry, so the company called Genomatica, they\u2019ve actually programmed microbes to produce fully biodegradable plastics, and so now they can produce biodegradable plastics at scale to replace things like plastic bags. And then you look at the textile industry as a whole. We can program yeast to produce leather or even spider silk, so there\u2019s a whole range of technologies that are really exciting.<\/p>\n<p><strong>Host: So, this research is incredibly ambitious. It takes a lot of brains, a lot of expertise. We\u2019ll talk about partners in a minute. But I want to talk right now about the main project that you\u2019re working on. It\u2019s called Station B. So, we\u2019ve identified some of the problems inherent in programming biology as well as some of the sort of individual trial and error attempts to solve them. But this is a much more comprehensive run at this hill. Tell us all about Station B. What is it? How\u2019s it different? What\u2019s it going to do?<\/strong><\/p>\n<p>Andrew Phillips: So, Station B is really motivated by all of the applications that I just talked about, right? And so, there\u2019s this tremendous potential, but yet there are these tremendous, you know, barriers to achieving that potential. And the one I mentioned is just the fact that programming biology is primarily done by trial and error. And so, you know, there are many aspects to that that we can try and address. So what Station B is aiming to do is develop a platform, a system, that will transform programming biology from what is currently a process of trial and error to something that\u2019s systematic and predictable. And that requires bringing together many different pieces of the puzzle. In programming biology, there\u2019s a sort of standard \u201cdesign, build, test, learn\u201d cycle. So, we\u2019re trying to combine these different stages of programming into an integrated platform. And in the design phase, we\u2019re developing biological programming languages and compilers that can take programs written in a language that people can understand and compile them down into DNA, you know, code, that living systems can execute. In the test phase, we\u2019re partnering with a company called Synthace. They actually specialize in lab automation. They\u2019re one of the leading lab automation companies. And what they\u2019re doing is developing device drivers and an infrastructure layer to actually make it much easier to program lab equipment, lab robots, to do experiments more systematically and reproducibly by digitally encoding those experiments as programs. And Synthace is actually built on top of Microsoft Azure Internet of Things technology. So, I\u2019ve got design. We\u2019ve got build. We\u2019ve got test. And in the learn phase, we\u2019re actually combining expertise in machine learning to analyze the data in order to learn models of how biological systems compute. So, we\u2019re sort of proposing models, using machine learning to actually refine our hypotheses, and then storing that information, that knowledge, inside a knowledge base so that as we go around this \u201cdesign, build, test, learn\u201d cycle, we\u2019re actually getting better at understanding how to program biological systems. And so the key point here is to try and bring together these different technologies. And over the past decade, almost, we\u2019ve been working on individual methods, individual pieces, individual programming languages. And now with Station B, we\u2019re trying to bring together the individual methods we\u2019ve been developing and, you know, some of the breakthroughs that we\u2019ve made, into this integrated system that will help our partners and collaborators become better at programming biological systems.<\/p>\n<p><strong>Host: So where is this now? It\u2019s very much still in the research phase, yeah?<\/strong><\/p>\n<p>Andrew Phillips: That\u2019s right. We do have a research prototype of this platform that we\u2019ve developed. The next phase now is to actually work very closely with a selected number of partners in order to develop and apply this platform to specific challenges.<\/p>\n<p><strong>Host: So, let\u2019s talk about those partners for a minute. You\u2019ve got them across industry and academia. Who are you working with, and what kinds of things might we expect to see?<\/strong><\/p>\n<p>Andrew Phillips: Well, we continue to work with many university collaborators around the world on a range of specific research projects. But really, the first university collaboration involving Station B as a platform is with Princeton. There, we\u2019re working with Professor Bonnie Bassler, head of the Molecular Biology Department, and also Professor Ned Wingreen, a biophysicist by training, on understanding the mechanisms of biofilm formation. So, biofilms are essentially surface-associated colonies of bacteria, and they actually kill as many people as cancer, and they are one of the leading causes of microbial infection worldwide and also an important cause of antibiotic resistance, which was recently highlighted by the World Health Organization as a growing crisis that we cannot ignore. So, what we\u2019re trying to do is use the Station B platform to understand how biofilms form. What are the mechanisms that they use? And the platform, as I say, will combine programming languages and analysis methods to allow us to program microbial systems, perturb these microbial systems, measure the effects of those perturbations and try and reverse engineer how bacteria communicate and how they interact in order to form these biofilms. And then, by understanding the mechanisms of formation, we can seek to disrupt these biofilms, and potentially, hopefully in the future, that would give rise to new forms of treatment.<\/p>\n<p><strong>Host: And by biofilm, you mean slime.<\/strong><\/p>\n<p>Andrew Phillips: Yeah, that\u2019s right.<\/p>\n<p><strong>Host: Well, I mean, let\u2019s get real. So that\u2019s fascinating, because one of the things that we think about when we think about what kills people and what\u2019s bad, is disease. But where does the disease come from? So that\u2019s what you\u2019re addressing, right, is if we can get to the source, we can control more of it?<\/strong><\/p>\n<p>Andrew Phillips: That\u2019s right. I mean, for many years, you know, the pharmaceutical industry has almost been forced to do things, again, by trial and error. There\u2019s a disease and we have a hunch as to what molecules we want to target. And then, you know, pharmaceutical companies and researchers will just test the whole range of random compounds, see which ones stick, and then maybe put those in mice and then maybe eventually put them in people, without often knowing how these drugs are working. But now, as treatments become more sophisticated and as we get better at treating disease, it\u2019s becoming increasingly important to understand how the treatments work, and that requires an understanding of how the disease or the pathogen works.<\/p>\n<p><strong>Host: This is so cool. Because if you look at science over the eons, it\u2019s been, what happens if I put this with that? And your efforts here are to codify and shrink down that process of trial and error by using computer science.<\/strong><\/p>\n<p>Andrew Phillips: That\u2019s right. And I do want to emphasize, you know there\u2019s a whole field, and there are many people around the world working on this, and we\u2019re, you know, a part of that field. You know, at Microsoft, we do have expertise, and many years of research and breakthroughs in biological programming languages, compilers, machine learning methods but we\u2019re part of this growing field that\u2019s really trying to solve some of the most important challenges facing humanity.<\/p>\n<p><strong>Host: So, who are some other partners that you\u2019re working with in Station B, and what are you working on with them?<\/strong><\/p>\n<p>Andrew Phillips: So, our main other partner is Oxford BioMedica. And, as I mentioned briefly before, they essentially have developed technology to reprogram a patient\u2019s own immune cells to target specific cancers. And they are the first company, together with Novartis, to actually have FDA approval for this type of treatment. And in clinical trials, 80 percent of patients who actually had no hope of surviving, many of them had had a bone marrow transplant or had gone through chemotherapy, 80 percent of these patients, when they received this treatment, were in complete remission. And the treatment has also been approved by the NHS, National Health Service, in the UK, but at a cost of \u00a3282,000 pounds per patient. And so, these treatments are really expensive. And part of our collaboration with Oxford BioMedica is to try and work with them to improve the ways in which these treatments are produced and try, by understanding how the cells are functioning, how the cells are producing the treatments, to actually bring down the costs, but also to help with the development, in the future, of new treatments. There\u2019s a whole range of diseases, including diseases like Parkinson\u2019s disease and others, which could benefit from this type of technology that Oxford BioMedica and others in the field are developing. And so, we\u2019ve just started a collaboration with Oxford BioMedica to help improve the way that these treatments are produced and to look at ways of producing new treatments as well. What we\u2019re doing is working with this company in particular to try and help improve their existing technology and bring down the costs and allow them to develop new technologies, which in turn will be subject to the rules and regulations of the industry. Oxford BioMedica, their treatment is saving lives today. And with our Station B platform, we are looking forward to working closely with them to help save more lives tomorrow.<\/p>\n<p>(music plays)<\/p>\n<p><strong>Host: All right, Andrew, with all the promising futures, including winning the war on slime, your research is, at its core, about altering biology via computer coding. What could possibly go wrong?<\/strong><\/p>\n<p>Andrew Phillips: Yeah, good question. Well, as I said before, we are very careful about who we work with. And the two main partners we\u2019re working with for Station B, Princeton and Oxford BioMedica, they are subject to, you know, very stringent regulations that they abide by. And they\u2019ve been doing this work for many years. And, as I say, as new treatments are developed, then those treatments will go through the same, or even more rigorous, approval processes. So, we\u2019re really working with the right partners to try to help them before more productive. So that\u2019s one point. What\u2019s also very encouraging is that governments are taking this technology very seriously, and they\u2019re the ones who are setting the agenda and there\u2019ve been counsels appointed by various governments to study synthetic biology and the desire to program biology more effectively. And this situation is constantly being monitored. And as regulations are produced, then our partners will abide by those. So yeah, we have to be very careful in that respect.<\/p>\n<p><strong>Host: Well, let me push in a little bit there, because we have so many best-case scenarios in front of us on how this technology could be really helpful in our lives. But I can think of several, if not numerous, outcomes that might fall in the dystopian bucket of technical advance. So even as you think about how governments and agencies can try to regulate this, is there anything that keeps you up at night?<\/strong><\/p>\n<p>Andrew Phillips: What keeps me up at night right now is all of these challenges we face as a species, you know, sustainability and disease and environmental pollution\u2026 That\u2019s what currently keeps me up at night. And I see this technology as a way, as I mentioned in many of the applications I talked about, as a way to solve so many of these challenges. There\u2019s also another issue, which is that, you know, what if we do nothing? So, nature itself, interestingly enough, is constantly evolving. Natural organisms are constantly mutating. Viruses are mutating. So, nature is producing new diseases, naturally, constantly. And we\u2019re seeing, in some cases, resistance to medicines like antibiotics that have saved hundreds of millions of lives. These systems are now becoming resistant to antibiotics, and so we need to find new treatments. And so, if we do nothing, there is a real danger that a global pandemic breaks out that nature has produced through random mutation that we are unable to treat because we don\u2019t understand how these systems work and we\u2019re not able to develop the treatment in time. Or as these existing treatments start to fail because nature, again, is mutating and smart and outcompeting us and going around our treatments. If we don\u2019t understand how to develop new treatments quickly enough then we\u2019re in real trouble. So, I think there\u2019s a real threat from nature itself. But there\u2019s another important issue as well, which is, you know, as I mentioned before about the drug industry traditionally doing things by trial and error, and now we see this new potential still being done by trial and error. It\u2019s going to be increasingly important to do things in a predictable way, to do things systematically, to be able to understand what we\u2019re doing. I think with computer models, programming languages, machine learning, being able to close that loop between models and experiments, we\u2019ll be able to predict, more and more accurately, the outcomes of the modifications we\u2019re making so that we can be very careful about not making the wrong modifications. And if we get better and better at counteracting the bioterrorist that is nature, which is constantly throwing things at us, we\u2019ll also get better and better at counteracting human endeavors which are trying to be malicious, because now we understand that if a random mutation happens or a deliberate mutation happens, we\u2019ll be able to counteract it. So, I think it\u2019s going to be really important to stay on top of this.<\/p>\n<p><strong>Host: Andrew, tell us about yourself and your academic background. You\u2019re originally from Barbados, West Indies. You went to Toulouse, France, and now you\u2019re in Cambridge, England. You\u2019ve had quite a journey. What got you started, and how did you end up at Microsoft Research?<\/strong><\/p>\n<p>Andrew Phillips: Okay, so I was always interested in robotics, engineering. I was fascinated by machines that people designed. And so, I studied engineering in Toulouse, France, and then I got really interested in programming. And so, I learned computer science in Cambridge, did a PhD at Imperial College, in London, and studied concurrent, parallel computer systems. So, programming languages for programming these parallel systems, the theory and also the implementation techniques. And there, while at Imperial, I met Luca Cardelli, a scientist at Microsoft Research at the time. And he was of a similar background but a leader in the field of concurrent programming languages. And he was applying these to study biological systems, which are massively concurrent, and I got fascinated by this. And so, I did an internship at Microsoft Research, and then I was hired by Stephen Emmott, who was leading a team at the intersection computer science and biology. And that\u2019s how I got started. Since then I\u2019ve been trying to develop methods from computer science but that are specific to biology. And there\u2019s been a lot of cross-fertilization there.<\/p>\n<p><strong>Host: So, did you actually come up with the programming language to translate from binary code to DNA code, as it were?<\/strong><\/p>\n<p>Andrew Phillips: Well, actually, I had an intern, very, very talented intern, back in 2009, Michael Pedersen. And we worked together on this very preliminary prototype of a programming language which he coded up and then we published a paper together and designed the language together. And since then, we\u2019ve sort of been evolving and extending the language, and more importantly, trying to bridge the gap between what you write on a computer and what gets executed in a cell, and making sure that that\u2019s more and more predictable. So, we started a long time ago. I think we still have a long way to go in the future, but we\u2019re making progress.<\/p>\n<p><strong>Host: Is anyone else using your language?<\/strong><\/p>\n<p>Andrew Phillips: So, we\u2019ve developed, actually, three main languages. One for programming systems at the molecular level, another at the genetic level, and a third at the network level. So far, we\u2019ve had most success at the molecular level, because it\u2019s much more predictable. This is sort of programming DNA systems to compute. And so yes, there are a number of people who have used that language. I\u2019ve also taught some courses at this international, genetically engineered machines competition on using our genetic programming language. So yeah, we do have people using our software, but we are actually very careful about who we collaborate with&#8230;<\/p>\n<p><strong>Host: Right.<\/strong><\/p>\n<p>Andrew Phillips: &#8230;and using the software mostly internally.<\/p>\n<p><strong>Host: Do you have names for the languages?<\/strong><\/p>\n<p>Andrew Phillips: Yeah, we have one, it\u2019s called Visual DSD, DNA Strand Displacement, another one is Visual GEC for Genetic Engineering of Cells, and the third is RAIN, Reasoning About Interaction Networks.<\/p>\n<p><strong>Host: I love that. So, every so often I get a researcher on the show who has such an interesting side quest that we have to go there. I\u2019m not even going to ask you about all the things you\u2019ve done like snowboarding, kite surfing, Chinese kickboxing, Thai boxing \u2013 you\u2019re just like this extreme guy. But you\u2019re a qualified ballroom dance instructor and you were a member of the Imperial College Dance Team while you were getting your PhD. So, I just\u2026 I have to know, how did you get involved with the Strictly Ballroom set?<\/strong><\/p>\n<p>Andrew Phillips: Okay, so how I actually got started was, I was sort of looking forward to my wedding and wanting to make sure that I did a good job on the first dance. So, I thought I would attend a couple of ballroom dancing classes. But then I got invited to audition, and it all took off from there. So, I was part of the university team. We used to travel around the country and compete with other universities. It was great fun. We used to have lessons, you know, and practice several times a week. And I really enjoyed it. And then, you know, because we had to do all of that, I thought I could, as well, take the exams that qualify you to be a ballroom dance instructor. And so, I did those. Sadly, I\u2019m not so much involved anymore. That was a long time ago.<\/p>\n<p><strong>Host: Yeah.<\/strong><\/p>\n<p>Andrew Phillips: Uh, my best dance was, the waltz and also the foxtrot. I really enjoyed it. I still do the odd salsa from time to time.<\/p>\n<p><strong>Host: Awesome. All right. As we close, I like to ask my guests to leave our listeners with some parting thoughts. So sometimes it\u2019s advice, sometimes it\u2019s wisdom, sometimes it\u2019s predicting the future. What would you say to aspiring researchers who might be interested in the field of computational biology? What are the big, open problems, and what kinds of people do we need to help solve them?<\/strong><\/p>\n<p>Andrew Phillips: Well, the first thing to notice is that it\u2019s really an interdisciplinary endeavor. So, we need mathematicians, computer scientists, people with expertise in machine learning, programming languages, lab automation, and of course, biologists, experimental biologists. So, I would say that if you\u2019re looking to get into this field, it\u2019s really important to at least understand the intersection of these different disciplines or a subset of these disciplines. Someone who can do biological experiments but understands the principles of, say, machine learning, could really help make some of these exciting breakthroughs at the intersection of the two fields. The other thing is that I really think that programming biology is going to transform many of the industries that are in existence today. I think it\u2019s a sort of an underpinning technology that will help transform medicine, food, energy, and build the foundations for a future bio economy that\u2019s based on sustainable technology. So, it\u2019s really going to be an exciting field, and I would encourage anyone with an interest to join.<\/p>\n<p><strong>Host: Come help us.<\/strong><\/p>\n<p>Andrew Phillips: Exactly.<\/p>\n<p><strong>Host: Andrew Phillips, thank you for coming on the show today, and sharing all the insights in programmable biology.<\/strong><\/p>\n<p>Andrew Phillips: My pleasure.<\/p>\n<p>(music plays)<\/p>\n<p>To learn more about Dr. Andrew Phillips and how researchers are using computer science techniques to program biological systems, visit <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/\">Microsoft.com\/research<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Episode 67, March 13, 2019 &#8211; Today, Dr. Phillips talks about the challenges and rewards inherent in reverse engineering biological systems to see how they perform information processing. He also explains what we can learn from stressed out bacteria, and tells us about Station B, a new end-to-end platform his team is working on that aims to reduce the trial and error nature of lab experiments and help scientists turn biological cells into super-factories that could solve some of the most challenging problems in medicine, agriculture, the environment and more.<\/p>\n","protected":false},"author":38022,"featured_media":572184,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"msr-url-field":"https:\/\/player.blubrry.com\/id\/42463458","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"footnotes":""},"categories":[240054],"tags":[],"research-area":[13553],"msr-region":[239178],"msr-event-type":[],"msr-locale":[268875],"msr-post-option":[],"msr-impact-theme":[],"msr-promo-type":[],"msr-podcast-series":[],"class_list":["post-572181","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-msr-podcast","msr-research-area-medical-health-genomics","msr-region-europe","msr-locale-en_us"],"msr_event_details":{"start":"","end":"","location":""},"podcast_url":"https:\/\/player.blubrry.com\/id\/42463458","podcast_episode":"","msr_research_lab":[199561],"msr_impact_theme":[],"related-publications":[],"related-downloads":[],"related-videos":[],"related-academic-programs":[],"related-groups":[],"related-projects":[544545],"related-events":[],"related-researchers":[],"msr_type":"Post","featured_image_thumbnail":"<img width=\"960\" height=\"540\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788.png\" class=\"img-object-cover\" alt=\"a man wearing glasses posing for the camera\" decoding=\"async\" loading=\"lazy\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788.png 1400w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-300x169.png 300w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-768x432.png 768w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-1024x576.png 1024w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-1066x600.png 1066w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-655x368.png 655w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/03\/Andrew-Phillips_POD_Site_02_2019_1400x788-343x193.png 343w\" sizes=\"(max-width: 960px) 100vw, 960px\" \/>","byline":"","formattedDate":"March 13, 2019","formattedExcerpt":"Episode 67, March 13, 2019 - Today, Dr. Phillips talks about the challenges and rewards inherent in reverse engineering biological systems to see how they perform information processing. He also explains what we can learn from stressed out bacteria, and tells us about Station B,&hellip;","locale":{"slug":"en_us","name":"English","native":"","english":"English"},"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/posts\/572181"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/users\/38022"}],"replies":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/comments?post=572181"}],"version-history":[{"count":16,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/posts\/572181\/revisions"}],"predecessor-version":[{"id":573177,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/posts\/572181\/revisions\/573177"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/572184"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=572181"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/categories?post=572181"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/tags?post=572181"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=572181"},{"taxonomy":"msr-region","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-region?post=572181"},{"taxonomy":"msr-event-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event-type?post=572181"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=572181"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=572181"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=572181"},{"taxonomy":"msr-promo-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-promo-type?post=572181"},{"taxonomy":"msr-podcast-series","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-podcast-series?post=572181"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}