Anu Acharya is founder and CEO of MapMyGenome India, a Hyderabad-based genomics company. GenomePatri, its flagship product, helps you discover your genetic predisposition – what diseases you are likely to get, for instance – by profiling your gene makeup. The vision: “Better Health through Technology”. Acharya, who is an alumnus of IIT Kharagpur and the University of Illinois and holds twin postgraduate degrees in Physics and MIS, talks to us about the data challenges in her exciting but nascent industry, and tells us how MapMyGenome is taking away the fear of health data revealing nasty truths that stop many of us from getting a simple blood test.
Edited excerpts:
The narrative around startups, especially technology startups, is that they are made or marred by data. What has been your experience with MapMyGenome? Have you witnessed the full force of data in your business?
So in our case, I think we are building the [culture of data] as we go along. One of the challenges in genomics is that there is less data available for the Indian population, which is why we started this.
Could you delve deeper on the kind of obstacles genomics faces because of shallow data? What’s the big-picture situation in the industry?
If you look at the genomics revolution, it has been only [fifteen-odd] years since the first [study of the] human genome was completed. So even globally, data is increasing, but it is not yet deep enough for every kind of disease. For instance when it comes to rare conditions, we don’t have enough data points. For single-mutation diseases, data availability is fine. But for diseases that involve multiple genes, those are the ones where you never know whether for a different population, different markers will make a difference. Even within a single population there might be minute differences. As we move forward, we have to transcend those differences and get more individualized data.
[Atlan] does a lot of work with publicly available data across sectors. The quality and accessibility of that data is frequently a challenge. What is your view of the state of public data in India? What role you do think private data companies can play in making things better?
With publicly available data, you have to often tally the same data across multiple sources to make sure you eliminate errors.
If we take a marker, for instance, we don’t just go by one publication. We look up multiple publications. I would say the quality is improving. But we have also built our own proprietary database which gives us a very high level of control and quality.
Right. It is not the lack of publicly available data but the need to triangulate and clean the data that is the biggest challenge.
Yes, absolutely.
Earlier, you spoke about individualized data. MapMyGenome deals with very sensitive personal health data. How do you make sure that all this data is tamper-proof and isn’t misused?
Absolutely, the security of our customers’ data is paramount for us. We use Biotracker, developed by Ocimum Biosolutions, another company I founded, to manage and track information at every step of the process, from creating a customer’s profile to collecting their sample, and finally processing the sample and reporting the findings. Biotracker is compliant with major international standards, including those recommended by the US Food and Drug Administration, and it is used in a lot of large pharma companies.
Once you create your profile and order a test on our website, we send you a sample collection kit which is securely brought to our testing center by our logistics partner. Then Biotracker delinks all personal information from your sample and creates an anonymized, barcode-based ID, and only this barcode ID is accessible to our personnel at the processing stage. All reports are machine-generated and password-protected, which means nobody can access them or tamper with them.
We are also very transparent about seeking consent. Under no circumstances are we selling your data, which says “Here’s Tanmoy’s genotype information.” Neither do we use any information for research purposes, such as large-scale population studies, unless you explicitly give us consent to do so.
One can imagine that building awareness and trust in a business like this requires a lot of continuous dialogue.
Yes, communication is critical. During the entire testing process, we send regular emails to the customer about confidentiality and other aspects of the process. We want to make sure that the customer knows what is happening with their sample. Even when we generate the final report, we clearly communicate that only the first page of the report contains the customer’s name. The technician who processes the sample has no access to the name. Only the counselor, who advises the customer once the report is generated, can see the name but for a short period of time. Maybe we can do more, add more layers to our communication as we grow.
Okay, let’s talk about another facet of data in your context. When I first heard about MapMyGenome, I thought, ‘Wow, this is a bit intimidating’. How do you make sure that people aren’t intimidated by what their gene data might end up revealing? How do you tell people, ‘Hey, this data isn’t meant to be scary! It is an exciting new way of discovering yourself”?
We want everyone to know that just because you are predisposed to a disease doesn’t necessarily mean you will get it. The process of taking away the fear starts right from our name. When we launched, we were called the Indian Genomics Company. Later, we decided to name our product GenomePatri [akin to janam patri or horoscope]. Our brand message is “Discover your true selfie,” because we are literally helping you take your truest photograph and look at it up close. Our new product, Medicamap, analyzes your drug-response profile based on your genetic makeup. The tag line for that is “Understand the Why”. The idea is to say that we don’t just give you some raw information about which drugs work for you and which won’t, but we help you understand why this happens.
Also, if you look through your reports, the first thing you will notice is that we talk a lot about the individual’s traits—caffeine consumption, athletic performance, are you a warrior or a worrier, do you learn from feedback, etc. This kind of information may not be directly relevant from a medical point of view, but they tell you who you really are. You relate to it much more easily. We follow this up with information about likely diseases, etc.
Finally, we have our counselors who explain the data to the customer. If a layperson looks at the numbers in their report in isolation, they may not be able to make sense of it. If your report says you are twice as likely as others to develop a certain condition, what does that really mean? What can you do to reduce the risk? That’s where a professional genetic counselor can help.
I like to say we are dealing with an equation with a variable and a constant. Your genes are constant, but what you do with your genetic information is a variable. To determine the variable, you first need to know the constant.
Coming back to the role of data in startups, entrepreneurs are known to be intuitive people. How do you balance data-driven decision making with intuition?
Listening to data is very important because your intuition isn’t always correct. I believe in starting with a hypothesis and then testing it with data, rather than letting data drive your hypothesis.
That’s where entrepreneurs are different [from large companies]. They have a lot of time, so they start by amassing a lot of data. We don’t have time on our side. So often, we start with gut-based solutions to a problem, and then support our decision with the right data. If we are then proven wrong, we abandon the hypothesis and move on to the next one.
Every entrepreneur has a favorite story about a time when they disregarded the wisdom of data and took a gut-based punt. What’s yours?
I am sure there are hundreds of such instances, but one story that I remember well happened in 2005, at Ocimum. We were a small company, not even a million dollars in revenue. We were in the process of acquiring a customer who would have given our revenue a big boost. Suddenly, they went through some financial trouble, and one day I got a call from their CEO asking whether we would be interested in buying their company. It was a strange shift, from pitching for their business to being in a position to buy them out. On that phone call itself I said, ‘Sure, that sounds great’. That one decision had a big impact on the company. It’s not the way small companies make such decisions, but I did it. On the other hand, because we were a small company, we could jump at that opportunity. A larger company would have taken much, much longer. Though I should tell you, in the end the numbers also made sense!
Debjani Ghosh, former head of Intel India, told us that the biggest impact of data-driven decision making in corporate cultures is that bosses have to start accepting that they might be wrong. What do you feel will be the most important impact of data on society?
Today young people already believe more in the power of data than anything else when it comes to making life decisions. The point is, there is going to be a time when ‘data-driven thinking’ won’t be a separate exercise. It will be integrated with, and change, everyday experiences.
The challenge will be to make sure that all the training and learning that comes with it is personalized enough. We need to separate the insights from the data pertaining to an individual to that pertaining to larger groups.
Data is also changing fundamental mindsets. In the medical industry, there was earlier this skepticism that point-of-care devices cannot be as accurate as diagnosis by a human physician.
But today we have advanced to a point where you can no longer reject [machine-generated] data. For instance, when we do your genome sequence, that data is indisputable. It’s your unique ID. We are not making it up. Of course, we still need human experts to work through that data, flag exceptions, etc. There may be differences based on how humans interpret the data, but the base data itself will always remain the same.
Photo credits: MapMyGenome India
