(06:12)[DAMIEN BLENKINSOPP]: Michael, great to have you on the show.

[MICHAEL NOVA]: Thank you, it’s my pleasure.

[DAMIEN BLENKINSOPP]: How did you first get into the area of genomics, and now it’s personalized medicine, but was there an evolution towards that? When did this first start for you?

[MICHAEL NOVA]: I was a research associate at the Salk institute a while back in a Nobel Prize winner’s laboratory – his name was Roger Guillemin. It was a very large laboratory; it had a lot of different technologies and scientists that were involved with it, as you can imagine.

The overall function of the laboratory was to study growth factors, and so we were studying everything about growth factors. We were studying how the proteins worked, tissue culture, how they interacted with each other, the DNA and RNA genetics of these growth factors, everything you could think of.

[DAMIEN BLENKINSOPP]: When you say growth factors, what exactly would that be for?

[MICHAEL NOVA]: Things like human growth hormone and thyroid releasing hormone and corticotropin-releasing factor, every kind of growth factor.

[DAMIEN BLENKINSOPP]: Okay. Things that stimulate growth in the human body?

[MICHAEL NOVA]: Yeah, in one way or another. He got the Nobel Prize for the first person to isolate TRF, which was a growth factor that was released in the hypothalamus. A signal that is released in the hypothalamus goes to the pituitary and then turns on all these thyroid hormones. That’s what he got it for, and so we were just peeling back the onion on a lot of different growth factors using different technologies.

I got into genetics there and then I started a couple of companies and took one public in the biotech area. We’ve almost used genetics as part of the technology, but it’s only been recently when we started (with Jim Plante, the founder of Pathway Genomics), we put a team together to really go after genetics as a solution for patients, and using genetics and genomics, I guess, as a solution for patients, and also physicians, for risk assessment or to give them insight into personal issues and to try and take some action against it.

I think it’s really been within the last ten years that the technology has been inexpensive enough that we could even try to use it directly for patients.

[DAMIEN BLENKINSOPP]: Great. First of all, I think a lot of people have heard of 23andMe, but they haven’t necessarily heard of Pathway Genomics, so could you give us a comparison of how the technologies compare and how the service is different? I know Pathway Genomics kind of evolved over time, so potentially a bit of that back story would be helpful too.

[MICHAEL NOVA]: Sure. First of all, the major difference is we have our own laboratory; 23andMe doesn’t. We have a big laboratory staff and scientific staff and curators and all that. All the tests come back to our laboratory and we do the DNA isolation and we do the reading of the mutations on different types of machines and then develop a report that goes back to the physician, which is the second difference: we’re only a physician’s ordered test; we’re not direct to consumer. So there has to be a physician in the loop or some kind of health provider in the loop, certainly on the ordering side, but also on the interpretation of the test.

All our tests are covered by insurance in the United States – that’s a third differentiator. We sell our tests in 44 different countries…

[DAMIEN BLENKINSOPP]: So just on the insurance angle; I understand it that you’re targeting a much smaller range of genetics, and basically you’re targeting specific arrays of things that you want to look at, like pharmacogenetics and other areas of the human body, whereas 23andMe is very, very broad in terms of what they look for?

[MICHAEL NOVA]: Yeah that was going to be my fourth!

[DAMIEN BLENKINSOPP]: Oh sorry.

[MICHAEL NOVA]: You took the wind out of my sails with that one, that was going to be the fourth big differentiator. We offer, like you said, panels of genes. We have a test for fit, nutrition and exercise, which only covers those two elements and then some eating behaviors and some metabolism.

Then we have another test for pharmacogenetics, like you mentioned. And one which is specifically for psychiatric, and another one that’s specifically for pain medications. Then we have a cardiovascular test, a cardiovascular risk, which also has some diet and exercise components in it.

So we have about 12 different product lines, 12 different types of tests, including BRCA. We do whole genome sequencing or next generation sequencing for the entire BRCA gene, if you know that gene. It’s the one that is prominent in certain ethnic groups for hereditary breast cancer. It’s the same gene that Angelina Jolie had. So we test for that as well.

We’re the only comprehensive genetic testing company that has health and wellness products all the way to hardcore next generation sequencing products for risk assessment for things like breast cancer.

A new thing that’s coming is we have an alliance with IBM, who’s an equity partner, and we’re building a mobile application that will basically put an artificial intelligence super computer in a handset to help with managing patient information and giving recommendations back directly to the user. That will be a direct to consumer type of product, but at this point we don’t sell any of our genetic tests direct to consumer.

[DAMIEN BLENKINSOPP]: I’d like to take a little step back because 23andMe and you are really very different propositions. There’s also the technology and the accuracy of the tests, and you have a different price point as well. Whereas I think for 23andMe for the whole thing right now, it’s $99; per array, yours is roughly $199 per different panel. So why is that, what’s the difference in the technology and what you’re delivering?

[MICHAEL NOVA]: It depends on the genetic tests. We do Fluidigm assays for our smaller arrays of up to about 80 different genes; 23andMe doesn’t do that. They basically take an Illumina chip that’s got a certain number of markers on it and run that chip for their $99 test. We also have that chip-based technology and then we also have the sequencing technology, which 23andMe doesn’t have.

So we have, the sequencing technology is basically more expensive than the Fluidigm or TaqMan assays, which are probably the least expensive.

We run every different type of genetic testing in here, but some of our reports require more than one platform. Some of them require the Fluidigm platform plus either maybe a sequencing or plus an Illumina chip, so the cost varies on a particular report based number one, on the technology that we’re using – it could be more expensive to run that particular report.

Then the way we do the reports is also different. We have a physician that reviews the results, we have a dietician that reviews the results, we have all those people that are on staff that are patient and can access at any time, so there’s a little bit more cost that’s embedded into the test or tests, depending on which one the clinician orders from us.

[DAMIEN BLENKINSOPP]: Right. Are your tests 100% accurate, so we could run them one time and we’d know for sure which gene SNPs we have?

[MICHAEL NOVA]: Sure. We have our own laboratory and it’s CLIA certified, CAP certified, it’s New York State certified. We’re the only comprehensive genetic testing company that has a health and wellness panel that’s been certified by New York State, which is very difficult to get.

23andMe can’t sell in New York State; they can’t sell in certain countries because direct to consumers is illegal. It’s illegal in places like Brazil and Singapore.

Our accuracy, since we’re licensed by three or four different licensing bodies, they come in here and inspect us all the time, at least once a year on all of them. So, we have to be extremely accurate.

[DAMIEN BLENKINSOPP]: I guess what I’m getting at also is the chip set that 23andMe is using is pretty reliable but it’s not 100% accurate, as I understand it. So in the past when I’ve done tests – I’ve done the 23andMe and I’ve done some other more specific genetics tests – and the answers weren’t the same. As I understood it, it was related to the technology that 23andMe uses, which is very economical to get a lot of data – which is interesting, so look at a variety of risks – but if you want to actually get clinical based information where you’re going to make decisions, you should run with the sequencing technology that you’re using with your panels to be 100% certain. Or am I looking at the wrong things there?

[MICHAEL NOVA]: No, I think you’re right on one aspect or a couple of aspects of what you said. I think that for things like the BRCA test, which is a very serious type of genetic test, 23andMe only reports on a couple of variants on the BRCA mutations, whereas we run the entire sequence. So the doctors come to us for that particular test; they would not necessarily go to 23andMe, even though the mutations that they provide and the way they do it are probably accurate, but they, just by definition, miss stuff.

It doesn’t mean that their technology is bad, which it isn’t; it doesn’t mean that the way they run the Illumina chip is not sufficient. That’s not correct. For what they’re reporting on, it’s perfectly adequate.

[DAMIEN BLENKINSOPP]: So everything you get reported should be correct with their technology as well – the Illumina chip?

[MICHAEL NOVA]: Yeah and I think it’s a good company. 23andMe is a good company. There are good companies like us and 23andMe and some of the other ones – we’ve been at this for eight years or seven years; we know what we’re doing. We just happen to have our own laboratory and so we’re under a lot of different kinds of governance that 23andMe isn’t under.

(15:22)[DAMIEN BLENKINSOPP]: Do you use blood samples as well, or is it saliva samples?

[MICHAEL NOVA]: Sure, we can use blood, saliva…

[DAMIEN BLENKINSOPP]: Is there a difference in the quality, or is it exactly the same, it doesn’t really matter which one you use?

[MICHAEL NOVA]: Both samples have different pluses and minuses, but trying to get to the same endpoint you still have to conform to what the governing bodies and what the licensing groups want us to report on. So we don’t have any choice but to make them equal in the end – if you gave us a blood sample or a saliva sample. But the way we do each one… in some respects it’s harder to do saliva because there are more contaminants in it and whatever, but then it’s a much easier test. People don’t necessarily want to get needle stick all the time.

[DAMIEN BLENKINSOPP]: I guess I’m trying to understand like I had a blood test run through DNA sequencing and a couple of the SNPs were different compared to my 23andMe. What would be the cause of that or is it a mystery?

[MICHAEL NOVA]: We can’t do that necessarily. We would certainly have to report on the same SNPs in the report in the same way so I don’t know. It could be a number of different things.

23andMe, again, has been around for a long time and so I think the accuracy of their reports and what they’re reporting on is really good. It’s hard for me to make a kind of black or white decision on something like that.

[DAMIEN BLENKINSOPP]: No, no, I’m not talking black or white, I’m just curious if there was a technological basis or something like that.

[MICHAEL NOVA]: There might be.

[DAMIEN BLENKINSOPP]: Yeah, I just figured it was the slightly different configuration of the technology.

[MICHAEL NOVA]: I’ll give you a really good example here and I think people don’t realize it: If you went and got a SMAT panel or a CAM panel from one company, like LabCorp, or you went and got one and put in the same sample to Quest, there’s no question that there will be a little bit of difference in what each one of these things reported on, but just a tiny bit of difference. That doesn’t mean that they’re wrong – either of them.

People think that genetics is black and white and the laboratory results are exactly 100% supposed to be the same all the time; that’s not necessarily true. And then we don’t know a lot more about the genetics either: There are 25,000 different genes, and we probably know what about 10,000 of them actually really do, but then they have to work with each other and all this kind of stuff.

I think getting the information on the particular SNPs is not necessarily the hard part; the hard part is interpreting what it means and giving that information back to the patient.

[DAMIEN BLENKINSOPP]: So it may be just a different reporting basis, that’s what it sounds like.

[MICHAEL NOVA]: Yeah, it could be.

(17:46)[DAMIEN BLENKINSOPP]: Taking a little step back, because I know this is basically your area, what does a shift to personalized or precision medicine and health mean versus where we are currently in the world?

[MICHAEL NOVA]: As a physician, we’ve always kind of practiced personalized medicine. When somebody comes in and they’ve got some condition they’re worried about, we give them their medications or help based on them as a person. But now, we’ve got a lot more tools. There’s a lot more granularity in what we can actually see that might be affecting this individual or even preventing things from happening.

Genetics is just one of those tools. So there’s genetics, there’s epigenetics, there’s transcriptomics, there’s all these different types of technologies now that are becoming less and less expensive. They’re kind of getting weaved into the management, if you will, of patients, and that’s what doctors are doing, basically, with our reports.

Precision medicine is just another name for personalized medicine, but I think one of the reasons there’s a much bigger push for it now is that we’re really seeing some major advances in cancer-targeted therapies using genetics, we know cancer is a genetic disease, a molecular disease. We’re now starting to target individual mutations in these cancers to give better results.

We’re now getting a clearer understanding of things like obesity – there are 97 genes that are related to obesity – they’re all different metabolites. It’s not necessarily going to be one size fits all and now we just have technologies that are getting less and less expensive to weave in information for the physicians to make decisions on. That’s where it’s at right now.

This is going to be an ongoing process forever; there’s going to be some sort of genetics or -omics or precision medicine technology that we’ll be able to use to really personalize individual therapies or prevention regimes or whatever you want to call it.

b>[DAMIEN BLENKINSOPP]: I guess one of the things about personalized is, if we take a comparison: If you have a cough today, you’re given the same drug no matter who you are; but in the future – and you have a panel which is pharmacokinetics – you could look at the impact of the drug on you – depending on your genes, drugs have a different impact. So it’s taking it up to a much more personalized level than what is possible today by just looking at someone.

In some cases, maybe you’ll see they’re different and maybe have got some blood test that is slightly different, but the genetics adds another layer of personalization.

[MICHAEL NOVA]: This is standard knowledge in the industry that anywhere between 40 and 50% of all drugs that are prescribed fail for the user, and especially the first time around. That’s a huge number.
mu
If we can add some way of tailoring those drugs – maybe you take this antidepressant instead of that antidepressant or you take this cough drug versus some other cough drug because your liver is metabolizing it different based on your genetics – you’re more likely to get a much better result.

Again, that is certainly where everything is headed in this whole precision medicine area.

(20:40)[DAMIEN BLENKINSOPP]: Great. So I also just wanted to talk to you – your tests are insured compared to the other ones, so I guess that the extent of research done on the specific panels is quite deep to get to that level where now tests can be insured.

[MICHAEL NOVA]: Yeah it depends. I’ll just take Medicare as an example because they’re kind of the gatekeeper for insurance coverage and our tests are covered by Medicare. The way that Medicare does it now in the United States, it looks at a panel on a gene by gene basis, and some genes have more clearly defined outcomes and predictability than others. So, on a panel of 80 genes, they might only cover three or four of them, but that’s enough to cover the entire cost of the panel.

There are three big levels of gene coverage in America. There’s a) genes that are covered automatically, like methylenetetrahydrofolate and some of the genes for warfarin metabolism. These are covered automatically, it’s an automatic payment, and since the technology on the panel is cheap enough, at least for us, to get over the cost of doing just that one gene, whatever Medicare decides to pay us, we make enough money to cover the panel. That goes for all the other insurance companies too, whether it’s United Healthcare pays on certain things, Aetna pays on certain things. Some insurance companies don’t pay at all on genetics, one way or another, so it really is not just based on whether the data is good enough, but it’s also based on whether a certain insurance company thinks it’s relevant enough to pay for it.

[DAMIEN BLENKINSOPP]: Right, right. As you’re saying, only part of your panel will necessarily be covered by that, and then there’s other things you’ve added, which you feel are relevant too. How’d you make those decisions? What kind of level of research has to be done?

[MICHAEL NOVA]: Yeah. We have a very strong curation. We have, I think, 15 PhD level geneticists and genetic counselors, and myself and a number of MDs, and we basically go and we grind through the literature. We look for human clinical studies and see if the data is relevant enough or there’s enough human clinical studies to put the gene into the panel and then report on it. We can only report on what the human clinical studies tell us.

There are plenty of genes and plenty of studies out there that we never would report on because we don’t think it’s relative; we don’t think that the data is strong enough. So to give you an example, in our healthy weight and fit test – which is our most popular test by far – we rate the science level in the test.

A really good clinical scientific study, let’s say on thousands of patients, and it has to be replicated in the same ethnic group, showed the same results and hopefully over multiple times, then that gets four stars.

Then there are other studies that aren’t quite as well validated but we think that there’s relevance because it might only have been done in two or three clinical studies of 500 patients each, which isn’t necessary a thousand patients but it’s 500 and it does show the same phenotype or it does show the same direction for what the genetics is reporting on. That’s a pretty good study so that gets put in the test too.

[DAMIEN BLENKINSOPP]: Great. I was going to actually ask you which was your most popular test but you’ve already brought it up.

So in terms of what that test gives people, who’s asking for the test and in what conditions are physicians looking for this kind of test? Is it someone who’s had a recurrent obesity problem for a very long time? What are the kinds of conditions and what’s actionable about that information for the physician once he gets it?

[MICHAEL NOVA]: For that particular test, we have a lot of different types of physicians that order it. Some are obviously looking for weight management, weight control in their population. But we also have people that are diabetics that are trying to use it to control their sugar levels or hemoglobin A1c levels, so we have a whole group of anti-diabetic groups that are using the test.

We have cardiovascular groups: many cardiologists think that most cardiovascular disease can be prevented by diet and exercise changing, so we have a lot of cardiologists that order the test and try to put people on more balanced types of diets, more personalized types of diets. Not necessarily to lose weight but to cut down lipid levels and other things that cardiologists worry about.

Then we have performance groups: we have performance athletes, we have gyms like Equinox health clubs that order our tests for a lot of their gym members to either increase performance or put on muscle mass, depending on what exercise.

So basically we have a lot of different types of groups, not just one type of clinician or group that orders the test.

[DAMIEN BLENKINSOPP]: Great. Is there an example you could walk us through of one of the most actionable genes in that area which people look at?

[MICHAEL NOVA]: Well, on that particular test… or do you mean all our tests?

[DAMIEN BLENKINSOPP]: The most popular one, because you said this was the most popular, if there’s one specific gene that people watch out for more than others.

[MICHAEL NOVA]: I don’t think it’s one particular gene. There are about 80 genes that we report on and we chop up the test into basically seven different sections. One section has to do with what is the best diet for you if you’re trying to lose weight and we have four different diets. That’s based on 50 different genes and how they interact with each other. Then we give you a diet recommendation, whether it’s low-carb, low-fat, standard Mediterranean or balanced diet. All of our diets are based on Mediterranean, but some have lower carbohydrates than others; some have lower fats.

Then we also give diet plans along with. So that’s a very popular part of the test, that section.

Another popular part of the test is we have a behavioral section, which looks at things like eating disinhibition – “I can’t stop eating.” Those genes around “When I start eating, I can’t stop eating.” Those genes in your dopamine pathway. We look at sugars – “Do I have a sweet tooth? Do I tend to like sugars more?” So this whole behavioral section is a very popular chunk of the test as well.

Then we have a standard metabolism section – we look at things like do you have a tendency to have increased insulin? Do you have a tendency to have increased lipid levels? Those types of genes, and there are multiple genes in that section – 20 or 30 genes in that section, so that’s also a popular part of the test.

[DAMIEN BLENKINSOPP]: Right. One of the interesting scenarios I think is the diet, the high fat versus the low-carb and the low-fat. Because a lot of the dietary recommendations today, it’s basically which crowd do you want to go with? I’m with the low-carb crowd; I’m with the high fat crowd, high protein…

Some of the genes can be pretty significant in that area, like the APOE gene. Could you talk a little bit about that and how that influences your diet and whether fats are going to be good for you or are basically going to be problematic.

[MICHAEL NOVA]: Let’s go back and look at diets in general. Most people, if they got on a diet and it was less calories than they usually intake and they stayed on that diet for years, they would probably lose weight. But it’s very difficult to get people to do that for a number of reasons.

So what we try to do is we try to look at particular genetics around fat metabolism – and APOE is one of them, and PPARgamma, and even FTO and some of these other genes – and not only how you process fats but also how you taste things. You have bitter taste receptors that we look at.

People don’t eat things that they don’t like, so we try to tailor the diet based on a number of these big subsets, whether it’s how you metabolize lipids – and people that have two copies of the FTO gene, there’s no question that they have trouble metabolizing fat in a low carbohydrate diet than somebody that doesn’t have those. That gene has been very well characterized and is a known obesogenic gene along with MCR4. If you put those two genes together, people that have those two genes tend to be about ten pounds overweight than people that don’t have them.

So we take that information, then we go back and we design a diet that’s based around not only what your metabolism is but also what you potentially would like to eat and make it a diet that isn’t too rigorous, that you’ll never stay on, and then try to give you direct diet plans – basically what to eat, literally, on a daily basis: Breakfast, lunch and dinner, this is what you should eat.

Then we have diet specialists and nutritionists and exercise physiologists and all that stuff, that if you really need help with that kind of stuff, we have ways to get you that as well.

I guess what I’m getting at is we don’t like to look at genetics in a vacuum. It’s one part of a big puzzle, and the more pieces of the puzzle we can put together, the more success we have for personalizing things for the user. That seems to really work.

We have other 20,000 physicians in the US that are ordering our tests and they keep ordering it over and over again, along with our diet plans and whatever information we give them, and the results speak for themselves – they’ve shown that what they get out for their patients is really working.

(29:30)[DAMIEN BLENKINSOPP]: Can we just go back to a distinction that you made was that you’re not doing genetics, you’re more doing genomics, right – the interaction of all of the genes together? Is that what you mean by you were saying?

[MICHAEL NOVA]: That’s a little bit of a slicing that onion really thin.

[DAMIEN BLENKINSOPP]: So what is the approach? You’re saying that it’s not a good idea to look at just one specific gene on its own?

[MICHAEL NOVA]: Yeah, very few things are one gene and then you have something bad happen. Even then, even for things like BRCA, it’s still only a relative statistic. Even if you have BRCA and you’re Ashkenazi Jewish and have the mutations that are relevant, there is still only a 80% probability that you’ll end up having breast cancer. So that means there’s 20% that you wouldn’t have breast cancer.

So very few things are one gene, one bad outcome, fortunately. It’s usually multiple genes. Again, we talked about obesity – there’s at least 80 or 90 different genes that have something to do with making somebody obese. And how they all work together? That’s the gold nugget in all this business is how to figure out how they all work together.

[DAMIEN BLENKINSOPP]: The BRCA gene is interesting because they’re pretty extreme decisions, or as you say, very rational decisions, but a lot of people see it as an extreme decision that Angelina Jolie has taken and it’s been in the press and everything.

One factor into that is that there’s genetics versus epigenetics and how we approach genetics in practice when there’s potential for some epigenetic influence and where the gene’s not actually turned on or off, right? You don’t know which one it is – is it turned on or is it turned off? Were Angelina Jolie’s BRCA genes – were they turned on and, therefore, they did represent the risk?

So, just based on what you said there, you said there was an 80% chance – I don’t know if that was a real statistic with a certain BRCA gene, but would it be in that kind of order that they were looking at BRCA?

If you took your BRCA panel, even not looking at the epigenetic influence, is there an 80% chance that that risk really exists, without taking into account the epigenetic influences?

[MICHAEL NOVA]: Correct. And remember, BRCA was first isolated in the Ashkenazi Jewish population – that’s where it’s most relevant. Angelina Jolie had family members who had breast cancer. So her decision to have surgery was based not only that she was BRCA1 and BRCA2 positive but also the fact that her mother, I think, died of breast cancer, and she’s half Ashkenazi Jewish.

So there were a number of factors that went into her decision to have surgery, not only to have her mammaries resected but also to have her ovaries taken out. I think she went down that path as well because there’s an increased risk, potentially, for ovarian cancer, which is still a very serious disease.

So you have to take all the information in total. If there was no breast cancer in her family and she wasn’t part Ashkenazi Jewish, then there might be a reason to not potentially go down that path. But that’s up to her and her clinician to work that out.

That’s why we don’t think a test like that, which is a very serious test, should ever be direct to consumer. That, for us, is something that really needs some guidance along with trying to make decisions about that.

[DAMIEN BLENKINSOPP]: Right. Excellent. I think the epigenetics area – how do you approach working with your physicians and advising them?

Do you ask them to look at factors like you were just talking about hereditary? – what’s the situation with your parents, your grandparents; other things you can look at in conjunction with some of your tests in order to capture the epigenetics? – whether something’s actually taking place or not: Do you say, “You should run these blood tests if you get these genes, and thus you could make a better decision based on that,” or do you tend to keep it to the genetics themselves?

[MICHAEL NOVA]: We tend to keep it to the genetics at this point because epigenetics is fairly new. There’s not enough data – although I do totally believe in it – in a lot of respects for us to weave that in to the process of, “You’ve got this gene but it’s not turned off.” We can do that from a technology stand point, but there’s not enough clinical data to make really informed decisions around that.

[DAMIEN BLENKINSOPP]: Right. I was talking more, at this point, as you say, epigenetics is relatively new and it’s probably quite expensive at this point for you to be integrating that type of service.

[MICHAEL NOVA]: Those kinds of expression assays – although Illumina has a methylone chip, but I don’t think it’s a clinical grade thing – it’s definitely more expensive than the genetics.

(33:50)[DAMIEN BLENKINSOPP]: I was thinking more about metabolites and lipids and things like that. So for example, we were talking about the APOE, so if your cholesterol markers are off, that would be an indicator that that gene is switched on – correct?

[MICHAEL NOVA]: Yeah, something is definitely not working correctly or you’ve got something in your diet, also, that’s not the correct diet. Maybe you’re eating too much of X, you should be eating more of Y. So there’s, again, a number of different factors – genetics, epigenetics, proteomics, metabolomics.

The metabolomics and the proteomics and looking at lipid panels, those give you a snapshot, an immediate time of day, this is what your lipid level showed. What genetics does is give you a tendency towards where potentially the lipid levels in the long run will go if you don’t take certain actions doing certain things.

[DAMIEN BLENKINSOPP]: Yeah it does. I think the area of epigenetics is potentially very confusing to people because there is this aspect of genes potentially staying switched off. Say, for instance, exercise is an important mechanism for turning off – I’m not saying this is true – but the APOE gene, right?

[MICHAEL NOVA]: There’s been data that’s shown that FTO gene for obesity can be mitigated with certain exercise and diet regimes; those are known facts. There are starting to be really hardcore data around using the environment, and epigenetics is all around using the environment – what you do in your environment to turn genes on and off – and there is data around that.

That would be one example of something that in the near future we might end up reporting on. You can change how genes are expressed by something in the environment.

[DAMIEN BLENKINSOPP]: I’m sure at this stage it’s just at a discussion level with you and colleagues and other people that you know, but how far out do you think these kinds of things are, like being able to take the next step and understanding the epigenetic aspect of it and making decisions based on that as well as just the genetics?

[MICHAEL NOVA]: Epigenetics in some respects is even more complicated than the genetics because there are so many different things that can turn genes on and off: there are methylation patterns, there are acetylation patterns, there are phosphorylation patterns, which means molecules that actually bind the DNA, or histones or whatever, and modify things and turn genes on and off.

And then there are all the microRNAs. There’s thousands of different microRNAs, the junk matter in DNA that will turn genes on and off if they’re expressed or not. So it’s extraordinarily complicated!

(36:11)[DAMIEN BLENKINSOPP]: IBM is an equity partner in Pathway Genomics?

[MICHAEL NOVA]: Yes.

[DAMIEN BLENKINSOPP]: Right. I wanted to talk about Pathway [unclear 36:16] but I think it’s also relevant to what we’re discussing right now, it being so complex and everything. Are you looking at bioinformatics and things like that potentially in the future?

[MICHAEL NOVA]: See that’s what computers are really good at. They’re good at taking noise, basically. Whereas we would look at it and not come up with any pattern; a computer’s really good at making patterns out of things. They’re not necessarily sentient, but they’re really good at taking databases and huge amounts of information and then telling you that these two things are linked together – that’s what the information is. That’s basically what we’re starting to build with IBM.

We have a very strong bioinformatics group and engineering group, and this is an artificial intelligence. Basically, it’s the Watson artificial intelligence that can play chess and was on “Jeopardy!” the show in the United States. So we have to train it.

We like to say it’s a little bit like a dog: you train the dog by lobbing it a question and seeing what answer you get back and seeing if it’s relevant. 99% of the time to start with it’s not relevant, then you have to tell it why it’s no, and go back “It should be this instead of that.”

It’s a huge process to train, especially around health care, because there’s nothing that’s more data dense than health care data. It’s not just genetic data we’re interested in; we’re interested in your electronic health record, your lab results, your wearables – your Fitbit data and all that other stuff. We want to take all that information and then compare it to the standard of care that’s what’s going to be in the Watson engine, and then give you back a recommendation that’s really personalized.

If you asked a question like, “I’ve got a nose bleed” – if you have our mobile app Panorama – “I’ve got a nose bleed, what should I do?” you would get a different answer potentially than what I would because I’ve scanned all this different information about you and compared it to what is the standard of care, and since you’re a little bit different in this gene and your latest lab result is a little bit different over there and maybe you went for a run and fell on your face, all of those bits of information are really important in order to give you a decision or some sort of recommendation about what to do.

[DAMIEN BLENKINSOPP]: Right. That sounds incredibly ambitious.

[MICHAEL NOVA]: Sure.

[DAMIEN BLENKINSOPP]: But you are going to release something relatively soon, aren’t you, so what will that be when it comes out?

[MICHAEL NOVA]: We will have public beta, sometime September to October time frame this year. We’re going through trials right now with the alpha version.

Like you said, it’s a very complicated problem because it deals with a lot of different types of data, and then getting that data so Watson can understand it, which is a whole engineering task on its own, and then getting the right information into Watson – or IBM, the super computer, the artificial intelligence – and then getting the right and curated information in there so it has the state of art in what people are thinking in terms of health care.

So you’re right, it’s extremely ambitious, and we’re really, really excited about it.

[DAMIEN BLENKINSOPP]: Yeah I can imagine. It will be fun to use it when it comes out. Is it going to be sold through iTunes or something, how’s it going to work?

[MICHAEL NOVA]: Yeah, we’ll go through the iStore and all that, and whatever Android is.

(39:20) [DAMIEN BLENKINSOPP]: Okay great. One of the other things we touched on that I wanted to get a bit deeper into because I think a lot of people don’t realize how varied this is, is pharmacogenomics.

You have several panels; it’s quite extensive the number of panels, it seems, under that area, because you have mental health areas and other areas. Is it extremely varied the impact a drug can have on each and every person? Is this very common that drugs have very different impacts per person?

[MICHAEL NOVA]: I’ll start with the panel. We have two or three different panels for pharmacogenomics. One is what you mentioned, it’s a mental health panel that has things like anti-depressants, antipsychotics, mood elevators, 30 or 40 different drugs and they each are metabolized in your liver a little bit differently.

One drug is metabolized differently to another drug, and we look at those mutations in your liver enzymes – they’re called cytochromes.

Then there are also transport proteins that have variance in how the drug is transported from the blood into the cells. There are a couple of drugs in there that have different transport kinetics. Then there are some of them also that get excreted by your kidneys, and they have a little bit different kinetics.

So we put that whole panel together on mental health based on a lot of this genetic information, or the best that we could find. Doctors use it to try and start somebody out on a drug rather than guessing what this person should have, or they’ll change a drug based on the genetics because they’ll understand why this potential drug isn’t necessarily working.

Then we have other panels. We have a pain panel, which does the same kind of thing but around pain medications – the codeines, oxycodone, morphine, tramadol, things like that – they get metabolized differently.

[DAMIEN BLENKINSOPP]: When you say metabolized, it means processed by the liver?

[MICHAEL NOVA]: Yeah, processed by the liver. There’s also transporters and uptake and excretion that are a little bit different for some of these drugs. Again, we use that information on a broad panel of different genes to tailor what potentially would be better for somebody than something else.

That kind of data is getting better. The good thing about genetics in general is that the data just gets better and better; it doesn’t get worse. It’s not like cold fusion – it’s not going to go away. It’s just going to be integrated more and more into the practice and pharmacogenetics and, obviously, drug metabolism is a huge deal.

To give you a good example: in the Asian population, there’s a drug called carbamazepine and it’s used as an anticonvulsant. There are genes involved around the metabolism of carbamazepine that if you have these particular genes, you will probably have a very high likelihood of going into Stevens-Johnson Syndrome if you take carbamazepine, and that’s a very serious disease.

[DAMIEN BLENKINSOPP]: Stevens-Johnson Syndrome; could you just describe the effects of that because I don’t think it’s very common but it’s pretty horrific, right?

[MICHAEL NOVA]: Yeah, it’s an allergic reaction basically, an immune reaction against this particular drug and you can basically end up dying from it – you go into anaphylactic shock and your skin starts to slough off. It’s a really nasty way to go if you want to call it that way. But again, it’s not very common.

But it is common more in Asians, and so screening for carbamazepine is 100% done in South-East Asia, Taiwan, places like that that are still using the drug as part of an anticonvulsant regime. They won’t put anybody on it if that person comes up with that particular variant.

That’s a really good example of how using a genetic test will really literally dial out a lot of drugs or dial in a drug based on your genetics.

[DAMIEN BLENKINSOPP]: Right. Currently though today, it’s a little bit of a trial and error process if you see a physician. Even with antibiotics sometimes, unless you’ve had tests done, it’s trial and error. We’re working hopefully towards a place where there won’t be any of that trial and error, it will be eliminated over time by these kinds of tests.

With the caveat that epigenetics sometimes will have some influence, so it’s not 100% fallible. In terms of the pharmacogenomics, there’s still some potential that basically says “This drugs better than this one for you”. It’s not 100% fallible, correct?

[MICHAEL NOVA]: No. Again, what we try to do in the genetics business is report on what the literature tells us – period; that’s the bottom line – and is that result valid.

We know, in pharmacogenetics, that across all drugs, 40 to 50% of them fail when they’re first given, so that’s a huge problem. So, dialing in the right drug, even though it might not be 100% correct… although the Stevens-Johnson issue, with this particular gene and carbamazepine, is almost 100%, so there’s nobody in their right mind if they knew that that patient had those particular genes would put somebody on carbamazepine because that’s one of those issues that is almost really one gene, one effect – you just don’t do it!

[DAMIEN BLENKINSOPP]: Yeah, right, when the risk is so high. What other high risk ones are there? Is warfarin a big one?

[MICHAEL NOVA]: Yeah warfarin potentially could be a big one for a couple of reasons. A dosing of warfarin to begin with is a little bit difficult, you have to have really strong expertise in doing that. The way it’s done is it’s done over a period of time to figure out what your INR is and how you’re metabolizing it and then getting the right dose.

Warfarin is a serious compound; you don’t want to mess around with it. It’s basically rat poison and it’s a very serious anticoagulant, as are some of the other ones like Plavix. But if you can figure out initially which dose of warfarin is better for that individual based on its genetics, that’s a good thing.

Warfarin tends to be used when a problem arises, like potentially a stent or you’ve got some sort of other issue that needs anticoagulation so you need to put them on warfarin immediately. I think that having a point of care warfarin test for pharmacogenetics is probably the way that that is going to go. Nobody wants to sit around and wait for a day for some sort of genetic test to come back before they put them on a drug like warfarin if they need it immediately, if they’ve got an embolic stroke or something like that; you’re just going to do it anyway.

[DAMIEN BLENKINSOPP]: Right. That kind of information is helpful to have it already pre-done. That is why – it’s pre-empting the need for genetic data on you. In some cases it’s worthwhile doing, right? Cancer…

[MICHAEL NOVA]: Yeah, and then the holy grail in a certain period of time it will be 500 dollars or a thousand dollars to get a whole genome sequence of all your genes, all your DNA. Then everybody gets it done, insurance will probably pay for it, and it just gets put in your record at birth. That’s probably where it’s going.

If you look at the long-term goal of getting everybody genetically tested, that’s probably where it’s going to end up. Then you’ll just pull down the information when you need it – it’s already in your file, it’s in your electronic health record. Does this patient respond to carbamazepine? Does he respond badly to warfarin? You’ll just know that because you’ll just drop down the information electronically.

(46:04)[DAMIEN BLENKINSOPP]: Great, thanks for that. One other thing you mentioned, which I’m sure is going to be interesting to some people, is the athletics aspect and the performance there. Have you got any specific examples of genes you’re looking at and reporting that are useful for training or changing/optimizing there?

[MICHAEL NOVA]: Yeah, there’s a lot of genetics on VO2 max. Some people tend to have a tendency to have a higher VO2 max than other people based on their genetics. How do you use that information in order to tailor your workouts? Maybe you’re one of these people that has a low VO2 max, maybe you need to do more X exercise than somebody that has a tendency to have a higher VO2 max. So there are genes around that.

There are genes around power and endurance: some people tend to be more power people, which means that they respond better to power athletics or power sports than people that are endurance runners. There are some pretty famous genes in that power area – actin is one of them and ACE and some other genes.

Then there are genes around exercise and insulin response, exercise and sugar response. Our panel covers a lot of these and gives you a broad snapshot of what potentially would be a better type of exercise for you than somebody else.

[DAMIEN BLENKINSOPP]: Right. so the type suggestions would be resistance training versus endurance aerobics, cardiovascular kind of work – these kinds of recommendations?

[MICHAEL NOVA]: Yeah, and then a sophisticated personal coach – we use an Equinox personal coach – uses that information to tailor what types of exercise regimes, along with their diet, potentially would be better, you’d get more response around than something else.

(47:43)[DAMIEN BLENKINSOPP]: Great, thank you. Where would you recommend someone look to learn more about personalized genomics? Are there specific books or presentations of the subject that you know are good resources to learn more about this?

[MICHAEL NOVA]: I think we have a couple of them on our website, pathway.com. There’s a lot of them out there. The University of Utah has a very comprehensive genetics database.

If you really want to get down to hardcore genetics, all the genes are listed in certain databases such as GeneMed and NIH has a database of all the genetics and all the genes, all the variants and what they mean.

You can Google in “Genetics textbook” and there’ll be 50 of them that come up. Hospital groups like the Mayo Clinic has a really good genetics site, Harvard’s got a good one, Stanford and UCSF, they’ve all got really good information on those websites about genetics.

[DAMIEN BLENKINSOPP]: Great, great, great, thanks. How could people best connect with you and learn more about you and your work? Are you on Twitter or are you active anywhere else?

[MICHAEL NOVA]: Yeah, people lob in stuff to me all the time. I figure my email is usually the best way to get hold of me, or Twitter – we have a Twitter account from Pathway Genomics. A lot of information gets disseminated through the usual media outlets.

[DAMIEN BLENKINSOPP]: Alright, great. Is there anyone besides yourself you would recommend to learn more about this, for personalized approaches, whether it be pharmacogenomics or anything else?

[MICHAEL NOVA]: There’s a lot of academic groups, every major university has somebody that’s doing it. I could certainly give you a list of…

[DAMIEN BLENKINSOPP]: It sounds pretty broad. I don’t know if there’s anyone more in the populous base, potentially working with big companies like IBM or doing some similar work, potentially different in some areas to you that would be of interest?

[MICHAEL NOVA]: One person that’s been pounding the genetics drum bag for a long time has been Eric Topol, you’re probably familiar with him. He’s one of the leaders in personalized and putting the consumer in charge of his own health care. That’s basically what we’re trying to do here from a number of different angles.

(49:46)[DAMIEN BLENKINSOPP]: Great, excellent. A couple of questions now just on your own personal approach and view of body data; what kind of things have you had tracked for yourself, whether it’s genes or other biomarkers or fitness activity trackers? What kind of things do you track on your own biology?

[MICHAEL NOVA]: I’ve had my genome completely sequenced, so I know as much about my own genome as probably is available. So in that respect, I know what’s good for me. Then I’ve certainly changed around my diet a little bit and the types of exercise that I do based on what my genetics have shown me.

I do wear one of these Fitbit tracking gadgets, and there’s a lot of them; there’s a lot of different types. Then I’m going to for sure use Panorama, this health care app that we’re going to come out with, because it will be integrated into your cell phone. You type in “What shall I do for my exercise today?” and it will tell you, “Based on your genetics or lab results X, Y, Z, you should do this. You’ve already done a thousand steps, you should do this now. You can eat this. There’s a store around the corner, you can buy it there.”

There’s a whole bunch of different parameters that I think will be very, very useful in terms of tracking where you won’t know what’s really happening. I think that’s another thing that users will like about Panorama is there’s not going to be a lot of input; you don’t have to do a food log.

Users don’t want to do that kind of thing. We live in 140 character world!

[DAMIEN BLENKINSOPP]: Yeah, there’s a burden to collecting information.

[MICHAEL NOVA]: There’s a total burden. That’s a very good word to use. There’s a total burden and we’re trying to make it very easy for it to be done automatically, so you feel as though you almost have a guardian angel on your shoulder, in some respects.

[DAMIEN BLENKINSOPP]: Are you integrating it with existing sources of information or are you just making the app very easy to integrate? A bit like Evernote, which you can upload all sorts of things into it.

[MICHAEL NOVA]: Yeah, it will be both. You’ll be able to take what you want, or we’ll go out and find it. We’ll go get your Fitbit data, we’ll go get your electronic health record, we’ll go get whatever lab result, provided we get permission from you to do it, obviously. There’s consent that’s going to be involved in this whole thing.

We’ll try and make that, as you said, that burden or that bar really low. We’ll make it very easy for you to get a very inexpensive genetic test through the application.

[DAMIEN BLENKINSOPP]: So you’ll be able to buy a Pathway genetic test through the app and it will get integrated automatically?

[MICHAEL NOVA]: Yeah, or anybody else’s genetic test. Whether you’ve got 23andMe’s; we’ll integrate that information in there.

[DAMIEN BLENKINSOPP]: Great, great. Okay last question – I always ask this of everyone – what would be your recommendation to someone trying to use some data, any kind of data, to make better decisions about their health?

[MICHAEL NOVA]: Knowledge when it comes to preventing things from happening and to changing your behavior when it’s based on real science is a very powerful thing. We hear that all the time – “Oh, that’s why I didn’t like X or Y. Now I know it’s not all my fault. Now I can change it and stick to some potential diet regime with a lot more confidence and I’m going to get a better outcome.”

So for us, knowledge is power in order to change behavior, and that’s the name of the game for a lot of us is trying to change your behavior. Because you have a lot of power to be able to do that. Giving the consumer more information about themselves is a very powerful thing.

[DAMIEN BLENKINSOPP]: Right. It’s like once someone understands something more clearly, it gives them more clarity, it gives them more confidence; it makes it a lot easier to keep that behavior on board.

[MICHAEL NOVA]: Right.

[DAMIEN BLENKINSOPP]: Well Michael, thank you so much for your time today. I really enjoyed the chat.

[MICHAEL NOVA]: My pleasure.