On today’s episode, we char with Dr. Prasun Mishra. Dr. Mishra (Ex-Genentech, Ex-NCI, Ex-NIH) is the Founder and CEO of Agility Pharmaceuticals. He is a serial entrepreneur who founded his first company after graduating high school, and second company during his PhD.
Since then, he has accumulated a wealth of experienced by building, investing and advising numerous other companies. He is an investor/ co-founder/ board member of a few US based corporations focused on accelerating drug discovery & development, blockchain & digital health. Dr. Mishra is also founding president & CEO of American Association for Precision Medicine (AAPM) and is leading research efforts focused on preventing & curing chronic diseases; not only treating the sick but also providing knowledge/tools to individuals to live longer, healthier lives.
In this episode, you’ll learn:
- What is precision medicine?
- How is Big Data Analytics, AI and technology impacting the medical industry?
- Is the current process for drug discovery broken and if so, how can it be fixed?
- Will technology be at a point where research can be one step ahead of possible mutations in viruses or drug resistances?
- What does all this new technology mean for the skill requirements of health professionals?
We would also like to thank Atul Padha who made the introduction that allowed today’s interview to take place.
BOOKS AND RESOURCES
- Find Dr. Mishra LinkedIn
- Email Dr. Mishra
- American Association of Precision Medicine
CONNECT WITH SHAWN:
https://linktr.ee/ShawnflynnSV
- Shawn Flynn’s Twitter Account
- Shawn Flynn’s LinkedIn Account
- Silicon Valley LinkedIn Group Account
- Shawn Flynn’s Facebook Account
- Email Shawn@thesiliconvalleypodcast.com
Pre-intro (00:00)
You’re listening to the Silicon Valley podcast.
On today’s episode, we sit down with Dr. Mishra Prasun, who is the founder and CEO of Agility Pharmaceuticals. He’s a serial entrepreneur who founded his first company after graduating high school and second company during PhD. Since then, he’s accumulated a wealth of experience by building, investing and advising numerous other companies. He’s also the founder and president and CEO of the American Association for Precision Medicine AAPM and is leading in research efforts focused on preventing and curing chronic diseases.
We’ll talk about what is precision medicine? How is Big Data Analytics A.I. technology affecting the medical industry? Is the current process of drug discovery broken? And how could it be fixed? And in the future with technology, will researchers be ahead of possible mutations? And with the new technology, what new skills will be required by health professionals? This and much more on today’s episode of Silicon Valley.
Intro 01:01
Welcome to the Silicon Valley podcast with your host Shawn Flynn, who interviews famous entrepreneurs, venture capitalists and leaders in tech. Learn their secrets and see Tomorrow’s World Today.
Shawn Flynn 01:18
Dr. Prasun, thank you for taking the time today to be on Silicon Valley.
Mishra Prasun
Thank you, Shawn. Thank you for having me.
Shawn Flynn 01:26
Dr. Prasun, can you give us a little bit of background of your career up to this point?
Mishra Prasun
Yeah. So, Shawn, I come from a medical family and the philosophy that I’ll soon will live to help others live longer has been engraved in my genes and epigenes from very early on. So, with that philosophy, I took my career path, and a lot of decisions became very clear. So, my personal goal is how I can save as many lives as possible before I surrender mine. And with that, I trained myself in pharmacology. I studied at Rutgers University, where my whole career focus has been trying to bring new drugs to the clinic. New drugs and drug combinations to the clinic and help cancer patients. So, I spend some time at Rutgers. I was at the National Institute of Health by National Cancer Institute. I spent my majority of time there and then moved the Genentech Roche, where I enjoyed learning about drug discovery and contributing to the drug discovery process at a pharmaceutical giant and then moving on to start agility pharmaceuticals, which is a drug discovery, big data AI driven and technology driven drug discovery and development company.
Shawn Flynn 02:40
Your using big data AI all this technology for drug discovery, but can you give us a little bit of background about what big data is?
Mishra Prasun
Yeah, so that’s a really interesting question that you ask, Shawn. So, there is many definitions do it right? Big data. So, but the definition that I like is so the data sets and analytic techniques in applications that are so large. So, from terabytes to zettabytes to exabytes and complex from multiple omics, the sensor to social media data that they require advance and unique data storage, management analysis and visualization technologies. So, two key words so large and complex that they require advanced data storage management, analysis and visualization technology. So that is the basically definition of big data.
Shawn Flynn 03:29
So then how is big data used in health care?
Mishra Prasun
There are many sources of big data in healthcare and the applications. Big data can be used in various verticals and healthcare, including diagnostics. Which is data mining and analysis to identify causes of illness and preventive medicine, which is predictive analysis and data analysis of genetic lifestyle and social circumstances to prevent disease can be also utilize in precision medicine, which is leveraging in aggregate data to drive hyper personalized care and also in medical research. Big implications of big data driven medical and from ecological research to cure diseases and discover new treatment and medicines, and to reduce adverse drug events. So by harnessing big data and to spot medication errors and block potential adverse events that patient may and overall cost reduction by identification of value that drives better patient outcomes for long term savings and the last but not least population health, monitor big data to identify disease trends and health strategies based on demographic geography and socio economic sector. So, these are some of the areas where big data is being utilized right now.
Shawn Flynn 04:44
So then what are some of the challenges associated with big data analytics?
Mishra Prasun
Yes, so what I call big data since its share huge I call four Vs You know, there are problem that can be summarized in 4 Vs. Volume, Velocity, Variety, Validity. So, volume shared with anybody that surprise with this fact that 90 percent of the data that we have collected in the history of data is in just past two years. We are just collecting really, really massive amounts of data set. So that’s volume. So, there’s a sheer very high volume. Number two is velocity. So, data is being now literally being generated at the speed of light. It’s doubling every 40 months. So that’s just basically we are just collecting too much data. And then that’s on the case. The third is V is variety. So, coming from multiple sources. And this is most of times it’s unstructured. So, coming from the search data, social data, Web based behavior data, mobile data, crowd source data. So, there are all unstructured data. And that brings us to the fourth. The reliability of the data is often in question because they are not really annotated well. They’re unstructured. So, you know, so we are solving it by building algorithms that will deal both structured and unstructured data. And by generating more structured data with a lot of initiatives and studies that are being done to create more annotated structure data.
Shawn Flynn 06:08
Can you just go over the difference between structured and unstructured data?
Mishra Prasun
Yeah, so the structure data is when you have all the parameters are well-defined, like in a clinical trial. So, you have you’re running a clinical trial is a controlled clinical trial and then you have all the data Very well defined. Whereas unstructured data will be a data coming from your cell phone, coming from your social media, apps that are moderate in your heart rate. There are a lot of variables that are missing. They’re not controlled. It’s just the data is being collected. So, I think the best example would be we having a conversation whereas somebody recording our conversation and kind of structuring different paragraphs, topics and what was coming out of our conversation. I think that’s the best I could think of. But I think the clinical trial example is also a much easier way to describe it.
Shawn Flynn 06:57
So, what you’re trying to say is the structure data is unstructured, but with parameters, you could take unstructured data and use machine learning or that to make it structured based on what you’re looking for.
Mishra Prasun
And one thing is that meta-analysis, suppose there are so many clinical trial data, clinical trials that were conducted and we just collect all the data from all the clinical trials and then analyze it. But it then becomes a meta-analysis. But then that data is actually unstructured in certain way because there were all trials done and different. So, there’ll be a lot of noise actually that we have to account for. Whereas a structured data could be just one gigantic clinical study like a U.S. government does conducting all of us program where they are collecting lifestyle, clinical genomics data from many patients, actually. And so, this would be like a cohort of the structured data.
Shawn Flynn 07:48
It’s interesting, I didn’t know the U.S. government was working on that right now.
Mishra Prasun
There are more both private and public efforts in trying to create that structure data set, which will act as a reference point. Anybody is the reigning AI, what AI hungry for is the structure of data. And then once you train your AI, then we can feed information in it and then it can rinse the reference established right, and the new information can be fed. And then based on the reference, the new and the outputs can be achieved. So, this is all the efforts are to generate a structured data.
Shawn Flynn 08:20
You’d mentioned A.I. artificial intelligence. Can you talk a little bit about that, your definition, and also how does it compare to human intelligence?
Mishra Prasun
That has been a topic of shared debate, and I think Hollywood had a lot of fun with it actually. So as a data scientist, I can give you a data scientist perspective. So A.I. is that aspect of human intelligence modeled by computers. So, anything intelligent machines would do, is it closer to human intelligence? Would be is termed as artificial intelligence section. And machine learning is an aspect of artificial intelligence. It’s implementing aspects of A.I. to reprocessing data so that which we discuss structured, unstructured data. And the deep learning is an aspect of machine learning, which is modelling brain architectures with layers of individual classifiers, adding non-linearity. So, this is when pictures in your social media post knows who Shawn is. Also, the A.I. knows what is cat, what is dog, what is spaghetti. So that’s deep learning. But, you know, in medical field, we have use deep learning to identify diseases and analyze radiology data sets and trying to identify what are the abnormalities in those cans. We have utilized that in head field to our advantage.
Shawn Flynn 09:31
So then is the field of big data and A.I., are they converging?
Mishra Prasun
Yeah, I think. But before I go into, I think I didn’t finish your second part. How is it compare with the human intelligence? Right. This has been a topic of real interest in the media as well as in the scientific community. So, the topic was in the power page of science and news of eat and nature, Time. National Job [inaudible], Newsweek called the AI Doctor will see you now. Nature called the future offered AI as a future of workforce. Time had it on a cover, say instating artificial intelligence is the future of mankind. And National Geographic called the merging of Man and the machine. And the fourth is a bionic age. So, and fortune called humans are underrated when compared to robots and AI actually. So, this has been a very interesting debate. So, if you put the two systems together, the artificial intelligence and human intelligence, the robots actually are AI, they have worked at a fabulous speed. They operate 24/7. They have their own advantages and they are less biased actually and accurate. So, whatever we programmed them, they will do that. Whereas humans, energy efficient human body, especially human brain, is one of the most energy efficient organs in human body. And we are universal multitasking. Right. We could do. We don’t need to be programmed to do one thing where we can multitask. And we have complex moments, right? Moments we can. Robots, you will design it to do one task, but humans can do many things. You can tell them to bring coffee. Robots have to specially programmed to do that, actually. And that’s what we already have, billions of them. So, there are all these scaled up rates. So, we don’t have to worry about that problem. So, in conclusion, AI can carry out certain tasks related to monotonous judgments much better than humans. That’s what we are learning. However, humans are more energy efficient, multitasking and our already scaled up.
Shawn Flynn 11:21
I’m surprised that humans are more energy efficient.
Mishra Prasun
This is amazing. Human brain is, it uses the least amount of energy, actually. Of course, it’s the organ that utilizes the most amount of energy and out of any organs in your body. But overall, if you compare with the energy the machine uses versus what a human brain uses, it’s just a very fraction of if we learn from just how human body manages to do that. I think the whole science of robotics and, you know, will advance to another level.
Shawn Flynn 11:51
So, I guess the Matrix was actually more scientific based than I thought.
Mishra Prasun
Yeah, that’s that.
Shawn Flynn 12:00
So then going back to how is big data and A.I. converging right now?
Mishra Prasun
The whole field of data analytics began by. So, when we started e-commerce, you know, the transaction data and rack mounted infrastructure that we had, and we collected a lot of data as we did and in analytics. So this is when the computing extract transform load ETLs, it came into a picture where we were copying data from multiple sources and systems and bringing it into a different context in another source that actually a destination system which represented data entirely different from the source. We begin to analyze the data and we started treating reports charges. So, then that majority of that. So, this is the evolution of data. So, then the field next was moved to how to generate business intelligence from data. Right. And then so we started collecting so much data. Then we needed algorithms. It was beyond human abilities to analyze that. Right. Beyond Excel sheets, although a lot of people still use that. Then we started designing complex algorithms, algorithms that did that analytics for us and gave that. So, then we have that actually that evolution led to. Right. Now we have automation’s factory automation robot Siri, Alexa, and now they’re sitting all in our hands and our computers. These are user driven big data models for machine learning. Now, these machines are learning from our behavior. What they are doing is their next step is they’re generating a machine intelligence. So, they’re becoming intelligent about our behavior. And the goal is to reach a machine consciousness where machines will be conscious like in human being is towards our behavior and will much likely behave as individuals suggest to give you one example. The most famous example. So, you know, A.I. has defeated major champions and in games like Jeopardy! And an Alpha Go Alpha go was one of the products launched by Google. I think it defeated almost major champions, but one person still managed to defeat it. So, what the Google team did. They let Alpha goa learn to play the game of go itself by simply playing against itself. So, the algorithms started playing the game of goal by itself. Then the next generation of the product was launch and it achieved a superhuman performance. So, when it was brought against it, a previously published champion defeating Alpha Go the Alpha goes the newer version, defeated him 100, zero. That’s why the name Zero was coined. Alpha goes zero because no human was able to defeat that program. The goal is that the machines. Now the algorithm has played with itself and then learn to predict every move a human can make on that. So that’s the machine conscious. That’s a lot of people believe that we are still in machine intelligence stage. But I strongly believe that the Alpha Go Zero example, we have already reached the machine consciousness where machines are starting to build a consciousness towards our behaviors.
Shawn Flynn 14:58
So, Dr. Prasun, if right now, the machines are kind of a conscious level is artificial intelligence. Is that still even the right term or should it be called something else?
Mishra Prasun
Shawn, this has been a debate for like 50 years now since the field began. So, I’m a big advocate of augmented intelligence. Still, AI but not artificial intelligence, but augmented intelligence. So, in the field, in historically, the artificial intelligence, which was called it A.I. and intelligence augmentation, which is called IA or augmented intelligence. So, this has been a debate. So, the question is, are we designing these machines to have their own consciousness or operate by themselves or it’s just we are augmenting our own human intelligence and human with the cellphone. We all became smarter with the advent of G.P.S. We are now driving anywhere you want. You don’t have to do your evening homework or where you are going. We are becoming smarter. So, this is all augmented intelligence. So, I think the name should be augmented intelligence. And what I believe is a symbiosis of humans and machines is the way to go. Symbolically, Iron Man, that’s a future. But I think that would be. That’s what I personally believe. That is not artificial intelligence. And I think the community will realize and we will start calling it augmented intelligence in the future.
Shawn Flynn 16:15
So, let’s get back to medicine, what other areas of medicine will be affected by the advancements of big data analytics, A.I. and all technology?
Mishra Prasun
So, medicine, we are in the business of saving lives. It’s an AI machine learning is a boom to medical field. So, there are many fields that are currently being affected by advancements in A.I. and machine learning robotics. One of the fields is decision support in hospital monitoring. It’s one of the major areas that needs transformation. I’m hopeful that technology will do its job in advancing that. Also, medical imaging, medical imaging, we have seen great advancements in diagnostics and precision medicine is another area where I think that advancements and AI machine learning will play a role and of course, drug discovery. There is a lot of Drug discovery and development. There is a lot of intervention that we need in terms of data and analytics and how we can accelerate the field of drug discovery and development.
Shawn Flynn 17:14
I can’t wait to go into more information, all those different fields you’d mentioned. But before that, in the coming years of everything you mentioned, what excites you the most?
Mishra Prasun
I am in drug discovery and development so that I have a bias towards that. But there are other areas like decision support and hospital monitoring. So, when a patient goes to a doctor, provide a lot of electronic health records EHR which is called EHR market itself or so the revenue approach, or 30 billion. This was a data generated in 2017. So, this is a lot of activities happening in IBM acquired explorers. They also acquired Truven multi-billion-dollar advance EHR data analytics capabilities. Google was given access to healthcare data of up to 1.6 million missions in the UK. And we all know the Roche’s acquisition of EHR company flat. And also V.A. and Cerner. They reach into contract that was ten billion dollars. And Apple launched in 39 hospitals Apple Health Records. And this is something that was announced that Amazon launched as a medical transcription service. It’s called Transcribe Medical. So, this is like an Alexa. We have Alexa, the Amazon transcription service, listening to the conversation between physicians and patients and trying to transcribe records by itself. So, this intervention is needed because that will give physicians more time to talk positions. Physicians spend quality time and some other areas where now deep learning algorithms can diagnose diabetic retinopathy, skin cancer and other cancer lesions at par with physicians. Also, I’m very excited about advancements of A.I. in interpreting radiology reports, which is almost 91 percent accuracy right now, which is almost at par with the physician. So, these are some of the areas that I’m most excited about.
Shawn Flynn 18:59
With all these corporations now entering the hospitals with their technology, if something goes wrong, so, for example, with that transcription software, if it’s transcribed incorrectly, who’s going to get the lawsuit, the doctor or the software engineer? Who’s getting the blame?
Mishra Prasun
Well, that’s the question the field has, right, not just a medical field with all AI machine learning. So, these are something that, you know, we have to work with the regulators and the companies as a community to resolve. I think I can share with you some use cases and some of the learnings that field has done or all. When we are developing software, doctors and health care systems are heavily involved. So, in development and approval of sub-software and medical device. So, they have to literally sign off. Right. And so, once the AI software is functional, they will in insists on the end user agreement in the file and the developer against any claims misdiagnosis. So, this is the one way because, you know, it’s already been approved by the medical community and US FDA has already begun its approval process of software as a medical device. The definition is software that are intended to be used for one or more medical purposes and that perform these purposes without being part of a hardware medical device. So that’s as a software, as a medical device. So that is being regulated by FDA. There’s an interesting case in UK. So, the in UK Consumer Rights Act 2015, it contains an information relating to digital content, including software. Basically, every user may seek remedy if a software is not satisfactory, is not fit for a particular purpose, or does not meet the description. So, payers could recover damages by using a medical negligence claim. To AI, So, what I’m trying to say is AI can just be another tool used by hospitals and the air can be dealt under medical negligence team with AI merely being a tool used by hospital. So, these are just some parts of it. Overall, as a community, we will need to develop ethics guidelines and industry standards that will reflect benefits and risks and limitations of the AI related products.
Shawn Flynn 21:09
How does the skill set of the health professionals change with the impact of all this new software?
Mishra Prasun
There is ongoing debate. So, it is started with radiologist’s series of news reports that were published and they argued that radiologist who will not use AI machine learning will be replaced by radiologist who will. And as our field advances, I think what some have even said that physicians who will lose use a machine learning and area raw data or area evidence-based approaches would be replaced by those who will not. So, as we are becoming smarter, more intelligent by utilizing this technology advancements and technology, I think the healthcare is evolving to a next level. We all use a computer. If you’re a computer 10 years back and you have a computer today, there is amazing difference. That’s the same as a healthcare field. All the technologies, newer approaches, if physician are just as an example. That’s one motion that they may be replaced by those who do what I believe that if not replaced, but they will lose patients for sure. To those who will use these tools and I myself come from a background, you question regarding how it will impact the training. So, I come from pharmacologist. So, I used to run my experiments with hundreds and hundreds of plates myself. So, the goal was to play cancer cells and try to play different drugs in different concentrations and try to see what concentration kills and what drug response and what doesn’t respond. So, I perfected this whole thing with my own hands, and I could generate perfect curves. I was really proud of it and I trained even my students to do that. But then eventually what I realized that these all can be done by machines and this all can be automated. And that’s how I evolved. So, I moved to the National Cancer Institute and I utilized their high throughput robotic arms and robotic automation systems to do all those experiments for and ran one of the biggest national product libraries that out there. Then I started analyzing data and I focused doing something meaningful. Right. But then eventually I was able to even automate the data analysis part. So now I get the output that I get as data that is analyzed. So as a human now, I have advanced by just marrying technology. Right. I have advanced my abilities and, in a week,, I can run millions and millions of molecules and I can analyze millions of data points and I can make my decision what drug to move forward with or not. And as compared to 10 years back, or maybe more so, I think the whole field will evolve. I agree that we may need to train the next generation of workforce who will now do intelligence stuff and then who would be augmented by AI machine learning. And we all will go to the next level of the workforce and we all have to put those efforts to train that workforce.
Shawn Flynn 23:56
So, you talked a little bit about how new drugs were made. At least that’s kind of my impression there. What is the current drug discovery process and is there anything that could be changed in this process?
Mishra Prasun
The current drug discovery model is broken because it takes 2.6 billion dollars and 10 to 15 years to bring a drug to market. And this has resulted in a low supply and high demand scenario and that has led to prescription drug prices skyrocketing. The present drug discovery model is Ill-Equipped addressing biological warfare or disease outbreaks. So, I think what we have to really invest in, how we can streamline, what are the processes where we could use AI machine learning to accelerate those processes. So, there are certain areas where we could make a difference. There is a whole understanding. The biology part, AI machine learning can handle massive sets of data. There is also in medicine or chemistry. There is a lot can be done. Currently, the system is focused on creating their own molecules and testing. With AI machine learning intervention, we can screen millions and millions of compounds that will never synthesize. And also, that automation driven, high throughput biology is an answer sometimes in a lab. Humans make certain errors. But when we scale that biology, those errors will be all gone. There is no replacement for high throughput biology. So, I pointed out these problems and these are the solutions. And that’s how we’re trying to resolve.
Shawn Flynn 25:25
What about repurposing of drugs? I’ve heard of that term. Can you talk a little bit about that?
Mishra Prasun
The repurposing is interesting area. So basically, repurposing of drug means that, you know, there are already drugs that were approved for certain other indications. But can we use those drugs to treat some other diseases? And there are various ways to do that. So just to give you an example, there are ways to adjust to do a genomic profiling of yours. Suppose there is a cancer tumor. Right. And then we can see what other pathways that are affected. And based on that, there is a massive data sets of what pathways are drugs, different approved drugs hit. And we can match those pathways and then identify the drugs that would hit those pathways. And then those drugs are already approved. So, can we use those drugs as an experimental drug? And FDA has a path for the approval of those drugs. It’s much faster. That’s the repurposing, the drugs that are already approved because, you know, safety, efficacy, it already has been established because this is all a drug that has already been in humans. So, the usually the repurposing clinical trials are small in efficacy and have sufficient data to get an approval.
Shawn Flynn 26:35
So when a virus mutates and it’s resistant to the current drugs being used for it and doctors have to come up with new medicines to treat it, is there going to be a point in time where machine learning A.I. that they’re able to predict that mutations of these viruses before they happen.
Mishra Prasun
That’s the future, and we have to keep up with all those mutations and not only just in viruses, but in the targets in drug targets, you know, all these are moving targets. Both viruses and drug targets are known to accumulate mutations. So, we are trying to inhibit drug target. And what happens is that they try to somehow create a mutation that will knock off that inhibitor and let that target function again. There are a couple of ways they acquired resistance. That’s one way. And then the second way is kind of activating an alternate pathway. And that will let them survive that inhibition. So, you know, what we have done is we have collected massive sets of data on kinases and on enzymes and on viruses and learned about with the crystallization data. What are the changes that are driving in that binding site? What is the implication of structure, function, implication of that mutation and how we can inhibit that? So, can you use the example of the viruses that are mutating with HIV, Best example we have. We know that what are the recipes that are more prone to a mutation than we have designed certain drugs that will not only target the virus itself, but also we can give them in combination in case they will mutate. Right. So, all these learnings will in future will be applied to build that pipeline of drugs that we can build. So, the classic example was Gleevec. Gleevec was an anti-cancer drug that became like a foster child for precision medicine. And with that learning now, pharmaceutical industry is focused on developing. They start working on the backup around the time they’re trying to develop a new drug, because as the pharmaceutical industry, the industry has to be competitive enough to launch those drugs and be ahead of their competitors. So, this is now inbuild in the pharmaceutical drug discovery platform to come up with alternative drugs that target any mutations that will be in the target or in the virus.
Shawn Flynn 28:49
So then what are you working on right now, Dr. Prasun?
Mishra Prasun
At Agility, it’s a technology driven drug discovery company. And our goal is to accelerate the field of drug discovery by using technology, all of data analytics technique, big data, AI machine learning. So, the goal is to reduce the time and cost associated with drug discovery and development by half. So that will be a big contribution if we can bring that so that half time and half the cost. So that’s our major driving force. And what we have done is, so we are very good in building models of diseases and trying to understand biology of the models by sequencing by multi- omics. Yes, the genomics is not an answer. How we can understand the biology, you know, in a multi-omics function and integrate that in-house and publicly available data databases for drug discovery. So, once we know the biology and we know the target, we can use the deep medicinal chemistry ways-based approaches where we use a deep learning algorithm to mine for better drug candidate. So, from targets to better drug candidates and here we can use millions of compounds that were never even been synthesized by using AI machine learning. So, but in the field, it’s known that not all the targets that we identify are druggable. So, then we also have a third arm where we use high throughput biology. So, developing high in the advanced tools, technologies in large scale biologies to study the disease biology. So, basically using high throughput screening methods to identify upstream or downstream effectors that we can target if we cannot target the identified target. It’s undruggable. We could identify downstream effectors to target and still able to inhibit that progression of that disease. So, this three-pronged approach gives us an ability to inhibit any in a drug target that we identify. So that’s the in the future. And the goal is to move from past, which was symptom-based medicine to present, which is pattern-based medicine to future, which is algorithm-based medicine. So, what we are trying to do is we are trying to build integrated, personal, multi-omics signature and combining with imaging data that we have available and developing a personal molecular classifier. So once we have that personal model or classifier, if a new patient comes in, we can take the data of the patient and then we can feed it in the personal molecule classifier and we will know what drugs are, drug combinations work for that patient. And then we can deliver that promise through digital health to the physician. And the physician will deliver the promise of precision medicine, which is right drug at the right time. Right Patient at the right time. So, we started as a drug discovery company, very soon realized that we have to build a lot of algorithms and solutions to solve several problems in drug discovery and development process. So, we are now becoming also solutions company. So, you know, a combination. So that’s why we’re that’s how we trying to make a difference in this field.
Shawn Flynn 31:44
So, you’d said precision medicine. Can you go into more detail about that in your opinion of it?
Mishra Prasun
Yes, so, the precision medicine can be defined as an emerging approach or disease treatment and prevention that takes into account individual variability in genes, environment and lifestyle for each person. So that’s basically a definition of precision medicine, and the approach that allows physicians and researchers to predict more accurately which treatment and prevention strategies for particular disease will work in what group of people. And in contrast to one size fits all approach in which disease treatment and prevention strategies are developed for an average person. And you know that there is no such thing as an average person so with the less consideration to and the whole approaches is kind of all approach. And we have to change it with a precision medicine, which is the right drug, right patient, the right time accounts the individual variability in their genes, environment, lifestyle differences.
Shawn Flynn 32:42
So, I have a question on kind of individuals. Maybe this is really precision medicine, but those companies that collect your saliva for DNA tested, how valuable is that information and what can be done with it in the future?
Mishra Prasun
Basically, the company that are collecting saliva, they they’re doing either genomics right, or trying to get your information from your genome and match it to what is already known about that variation, differences that you may have. And then they are giving out that information to you. And FDA has intervened into the whole process. When, you know, somebody is trying to say that they’re trying to diagnose or treat a disease, they have to get an FDA approval. And there are certain tests that are approved by FDA. So that’s the majority of them are trying to just get information, Genetic variation information in your genome and that what is being scientifically known sometimes not all of those recommendations will come true. Like, you know, I have personally known that folks who had known to be prone to Alzheimer’s, the disease never fathomed in their lifetime, at least for now. So, you have to take everything I say with a grain of salt. And it’s every individual is different. And a disease is a complex set of events occurring together to hope that answers your question.
Shawn Flynn 33:57
Who’s going to have access to all the patient data? Who’s going to own the patient data in the future?
Mishra Prasun
That’s the evolution that we are going through now as a healthcare field of data is a problem. Data silos are a problem. And as a community, we’re really struggling with how to get access to good data sets, how to get access to structured data which is necessary to train AI so, the or all healthcare field is moving towards a patient centric model. So, if you ask me about in future, who will own their data? I think patients will own their data and what they will be able to do with that data. They will make the conscious decision of who they may or may not grant access to their data. And there will be also ways, in fact, there are certain companies right now. They also provide a way to monetize that. And other companies like pharmaceutical companies and other diagnostics companies, they will be able to take advantage of that. So, they will be able to identify that data and still be able to analyze without compromising patient privacy. So, yeah, I think in the future, patient will be the owner of the data.
Shawn Flynn 35:03
What about any ethical issues that could arise from precision medicine or any of the different advancements in technology we talked about?
Mishra Prasun
All these ethical issues related to precision medicine; one is being what I already spoke about, briefly touched upon was data privacy. Data privacy has been one of the major concerns. And genetic discrimination law was passed very early on to prevent Americans that will from this discrimination. So, as we enter the next era of the technology and data analytics driven medicine, we have to also make sure that we are protecting data privacy of a patient. And we had DNA identifying most of the data that is being used to train our AI machine learning algorithm. Then there is some social issue of what I talked about, the data silos. You know, there are sometimes so different institutions have their own data and data is has become a currency right now. So, everybody wants to hold onto that data. And there are very few platforms that would allow data sharing. And one of the focuses of American Association for Precision Medicine search is an organization that is a nonprofit organization that I also lead. And that’s one of the goals of the organization, is to build the contortions to solve this particular issue of data sharing.
Shawn Flynn 36:21
And you’d mentioned American Association for Precision Medicine. Can you talk a little bit about what that is?
Mishra Prasun
The American Association for Precision Medicine or in short, AAPM, its goal is to accelerate the field of precision medicine to research, education, communication and collaboration to foster new medical breakthroughs. And our goal is to bring the four piece of precision medicine together, the four pieces of being public health planners. That’s IE-government, providers, hospitals, physicians. That’s providers, payers in that insurance. Insurance companies and patients, most importantly of the four pieces together to bring together to facilitate the dialogue between these to achieve the goal of delivering improved outcomes in the field of Precision Medicine AAPM. So, we have several initiatives. And you know, you can read it on AAPM.health, which is at our Web site. We not only organized meetings, trainings, symposiums, conferences, but also through those interactions. We identify problem areas where we can add value and build solutions in partnership with the stakeholders to accelerate the field position medicine. And we’re in a planning of AAPM Innovation Center, which will serve as a catalyst in the digital transformation of the healthcare industry.
Shawn Flynn 37:35
So, you mentioned cancer right there. Should we expect any major breakthroughs in maybe 2020 or in those years to come?
Mishra Prasun
I think we all have to work as a community to catalyze those. We are losing so many loved ones to cancer and I personally lost my elder sister to pancreatic cancer. And when she was diagnosed, my research was focused on pancreatic cancer. I knew that how resistant this cancer is. The tumors are really rock solid. Nothing. No drug enters those. I knew that we couldn’t we will not be able to save her and eventually we lost her. You know, that has impacted me personally. I am now on this mission to see how we can work as a community to bring prevention and possibly a drug that will cure. And we have seen a lot of so the cure word in cancer field was not being used until the advent of cancer immunotherapy, where some of the patients have right now seen complete remissions. And they have been you know, they are surviving for often up to more than four or five, six years since the field has been out there. So, we are seeing complete remissions and complete cures in the medical field. And as a community, we are all there to catalyze that and that change to build the pipeline of drugs that will eradicate this disease that we call cancer.
Shawn Flynn 38:59
If anyone wants to find out more information about yourself, what you’re working on, what’s the best way to go about it?
Mishra Prasun
appm.health, is a Web site, and you can reach out to me directly by email. My personal email is JPrasunMishra@Gmail.com again it is JPrasunMishra@Gmail.com. Happy to answer any questions.
Shawn Flynn 39:19
Great. We’ll have those links in the show notes. And once again, Dr. Prasun. Thank you for spending the time today to be on Silicon Valley. I also want to thank a tool who is the gentleman that made the introduction to us originally who allowed this interview to happen. I’ll have his contact in the show notes as well. And stay tuned, everyone, for next week. We’re also going to have another exciting guest to bring you what’s happening here in Silicon Valley. Thank you.
Mishra Prasun
Thank you. Thank you, Shawn.
Outro 39:49
Thank you for listening to The Silicon Valley Podcast. To access our resources, visit us at TheSiliconValleyPodcast.com and follow our host on Twitter, Facebook, and LinkedIn @ShawnFlynnSV. This show is for entertainment purposes only and is licensed by The Investors Podcast Network. Before making any decisions, consult a professional.
