The Center for Data Innovation spoke to Oscar Flores, chief executive officer and co-founder of Made of Genes, a Barcelona-based consumer genomics company. Flores talked about the evolution of genomics from the groundbreaking science of the Human Genome Project at the start of the 21st century to a consumer service today, and about where he sees the field going in the future.
Nick Wallace: Consumer genomics is a pretty new industry, but it seems to be growing. How did you get involved in this?
Oscar Flores: I’m a computer engineer by training. I decided to apply that to the biomedical field and so did a PhD in biomedicine, and this is where I learned how to deal with genomic data. Genomics data is truly challenging in terms of data management and big data.
I was doing my PhD in a joint program in the Institute for Research in Biomedicine in Barcelona and the Barcelona Supercomputing Center, using high-performance computational techniques on biological data. And then I wanted to leave academia and go into industry, so I did an MBA. It was during this MBA that we had to come up with a business plan, so I asked my former boss about a pattern that there was in the Center, about cancer genomics. We started looking at this pattern and how to apply it to the market. We realized there are a lot of different problems—not ones that the pattern would solve, but problems with genomics data management. After two years of iterating, we founded Made of Genes.
At the beginning, we wanted to provide professional services to physicians and researchers—the kinds of people who would use genomics data. But then we realized this was too expensive. Public health systems in Europe cannot pay for everyone’s genomics. We realized that if we wanted to do something, it would have to be either the patients or their insurance companies that paid for the sequencing of genomic data.
Once we had that, there were problems regarding privacy, potential misuse of the information, what might happen if a hacker got into the systems and gained access to a lot of personal information, what insurance companies might do with this information, et cetera. So we started to design our business model focused on privacy, rather than mining the data—there are a lot of different companies focused on exploiting the data to reduce the price.
So we became a transactional service—like PayPal—where you store your genome in a safe way, and then every time you want to use it to access another service, you use our platform. We provide an interface where you are informed about which parts of your genome will be accessed, and why, and who will have access to it, and we provide a secure platform where your data can be analyzed.
So that was the value proposition for customers. But this also makes a lot of sense for companies—insurance companies, hospitals, even governments—to offer genomic services to their customers or patients or citizens. So on the business-to-customer side, we provide a marketplace model, where you only have to sequence your genome once and can then use it for as many genetic tests as you like, and then on the business-to-business side there is a licensing model, where they can use our platform to deliver personalized medical services to their customers.
Imagine a public health system: if you want to implement precision medicine in a population, every time a patient goes to their doctor, this information has to be included it the patient’s medical record and considered in the treatment provided. So this means the doctor might need to access the genome one, two, or three times per year. When you have to do that, you can’t do the traditional genetic testing that involves taking the sample, going to the laboratory, doing the biochemical processes required to get the DNA, et cetera—that’s too expensive.
So at this point, it makes a lot of sense to say, “let’s sequence the whole genome,” or the relevant parts of it, and store that. Nobody needs to access all of it at any one time, but it’s like having a personal encyclopedia that can be referred to when needed. It’s cheaper to do that than to do a DNA test every time the information is needed.
So far, there are no public health systems that do this, but there are initiatives in the United States and the United Kingdom where they’re looking at the viability of doing it. But these are very large and complex health systems—and the United States’ health system isn’t a public health system at all, and it’s too fragmented. So I think we’re more likely to see this happen in small countries first, and the larger countries later.
Wallace: It’s been about fourteen years since the human genome was first sequenced. What has changed since then to make your business a viable proposition?
Flores: The first human genome took thirteen years and $2.7 billion to sequence. Now you can have a whole human genome for $1,000, and you don’t need thirteen years, you can do it in 48 hours. This implies two things. One is that there’s more data, because it’s cheaper and quicker to collect. This doesn’t just apply to human genomes—there are major advances in agriculture thanks to plant and animal genomics—but it also means better research for human medicine, and better outcomes for healthcare. We’re seeing things now that were unimaginable ten years ago. For example, we can say now that by applying genetic testing, you can reduce by 70 percent the hospitalization rate related to drug dosage among people using anticoagulants. You can avoid 20 percent of breast cancers. You can test whether you’re a carrier of certain genetic diseases, and avoid passing it on to your children.
From a technological point of view, it would be pretty challenging to store genomic data without cloud storage. A whole human genome is about 600 gigabytes. Of course you can compress it and do all sorts of things to optimize it, but at some point you need to deal with 600 gigabytes of data. There are some estimates that if current trends continue, then by 2025 there will be more genomic data than there is data on YouTube today. Without cloud storage, every hospital would need farms of servers to store their patients’ genomic data, and they will not do that—they have too many other problems with IT infrastructure. But in the cloud, the applications are fast, we can store data virtually for free nowadays, and there’s no capex investment needed to access large, high-performance computational infrastructure. These are the enablers of consumer genomics.
Wallace: What kind of personalized services are already available to people who have sequenced their DNA, and which do you think are the most useful? Is there anything you’d like to do with this data, but can’t yet?
Flores: The usefulness is linked to the research that has been done in the relevant fields. I would say that 95 percent of genome sequencing in in the world so far has been for cancer research. So applications of genomics for cancer research are probably the most useful right now.
But in the future, I don’t think that’s where genomics will have the largest impact. There are things that we do daily that we wouldn’t if we knew we had genetic predispositions that amplify their risks. For example, diabetes has a very strong genetic component, so if you know you carry the gene that predisposes you to developing diabetes, you might take more care about what you eat.
Or take smokers. 80 percent of smokers don’t get cancer. But 20 percent do, and there are certain types of cancer where 95 percent of the people who get them are smokers. So if you know that you’re a carrier of a genetic variance that predisposes you to lung cancer, most probably you will not smoke, because the odds are stacked against you. On the other hand, people might gamble a bit more if they know when the odds are in their favor. For example, you can know if your baby is allergic to gluten or not—there was no way to do that before. You had to guess—say if you noticed baby vomited after eating cereals. With genetic tests, you can know for sure, and you can feed your baby gluten in the confidence that it isn’t harming him or her. At some point, the public health system, or insurance companies, will cover this, because these are savings.
There there are the curiosity uses, like for searching ancestral records. There’s even a social network that’s a kind of Tinder for genetics, so you can check genetic compatibility between different genomes.
There are other things that are still quite new and still haven’t been researched in much depth. For example, sports coaches could use genomics to personalize diet and training regimens for elite athletes. A lot of companies are looking at this field.
At some point, DNA sequencing will be like having your blood group on file: something that just needs to be there when you go to the doctor. It also has value in identification, like fingerprints. I can also think of some fun uses: imagine a video game where your character’s skills are based on your DNA.
Wallace: Sequencing one’s genome probably isn’t something most people would think about doing right now. Why should ordinary people be interested in this? What motivates your current customers?
Flores: There are two kinds of customers. There are those who do it for health reasons, such as somebody with a family history of breast cancer and who wants to try to prevent it happening to them. This is the largest group. The others do it out of curiosity; it’s something new and cool and they want to try it out, or they want to know where their ancestors came from.
These are the two main drivers. And of course, lots of people won’t fit into either category, and don’t want to know if they have a predisposition to some disease, they’re happier not knowing. If you talk about this with your friends, you’ll probably find half of them will say they never want to do anything like this, and the others will think it sounds cool and might be keen to try it out one day.
Gene sequencing costs around $1,000, and it isn’t like buying a smartphone. The satisfaction of having your genome is not as immediate as that of having the new iPhone and being able to play with new games et cetera. So I think the driver for this to become mainstream will be investment from the governments and health services and insurance companies. The health system in Spain isn’t sustainable anymore. We can sustain business as usual for the next ten years or so, but the population is becoming older, there are fewer people working, and health care needs to become more efficient to sustain this type of population. So at some point, genome sequencing may become part of long-term efficiency drives, because it saves money in the future. The change will come from regulators, not patients.
Wallace: Genetic data is of course valuable on its own, but I imagine that it could lead to a lot of new insights when integrated with other data sources, like fitness trackers or health records. Are you exploring any of these kinds of partnerships?
Flores: Yes, absolutely. We’re not planning to integrate with fitness tracking—that’s a very different market—but it makes a lot of sense to link genomic data with health care records. It makes a lot of sense to have both sets of information on hand when a doctor prescribes a given drug, because one of the things you can get from your genome is how fast you metabolize a drug. This is why I mentioned anticoagulants earlier: if a doctor prescribes an anticoagulant, and genomic analysis shows the patient is a slow metabolizer of this anticoagulant drug, instead of giving 600 milligrams, the doctor might give him 300 grams to reduce the risks. Genomic data is something that can be integrated with different databases and services, especially health records.