November 2016

Personal genomic data for the masses

By Craig Macpherson Editor and founder of DNA Testing Choice

Direct to consumer DNA Tests

It’s hard to put myself in the mind-set of someone who’s not yet encountered a ‘direct to consumer’ DNA test but I’ll try. They may have caught an episode of Jeremy Kyle where the term ‘DNA test’ is synonymous with ‘paternity test’. Or perhaps they’ll have seen an ad for 23andMe online or on a poster – it felt like 23andMe were plastering the trains with ads at one point... Maybe an Ancestry.com ad on TV will have caught their eye, or perhaps, unfortunately, they’ve had to discuss a DNA test with their GP for the purpose of a diagnosis or because a condition runs in the family. From my perspective, the public tends not to seek out their personal genomic data, but rather a few develop a desire for it as a side-effect of buying a DNA test to establish paternity, to trace ancestry, or to look more closely at their health.

This was reinforced when I last met Oxford Nanopore, one of the fastest growing manufacturers of devices to sequence DNA. Right now they have no plans to sell their services to the public (nor do Illumina for that matter), as there’s a perception that the public aren’t really interested in buying their own whole genome sequence – the highest quality genomic dataset publicly available. Take Genomics England, they’re on a mission to sequence 100,000 genomes in order to better diagnose and treat rare genetic conditions. But despite being set up in July 2013, they’ve only sequenced 14,000 genomes so far, far fewer than the genetics community expected. Not only that, but only one participant to date has asked Genomics England for a copy of their personal genomic data. When they were told they’d receive it on a portable hard-drive in unheard of file-types, they decided not to take Genomics England up on their offer.

For the few individuals who have bought a 23andMe test; the futurists, the enthusiasts, the errant genetic genealogists, and those seeking a sense of ‘control’ in their lives (‘control seekers’ being the most responsive customers according to 23andMe’s marketing department), they may feel they have their personal genomic data already. The difference between sequencing and chip analysis is lost on most, and when told their exome is all they’ll ever need, most customers take this at face value. From what I’ve seen, the public are a few leaps away from knowing what they’d want out of their personal genomic data, and you could argue that they may not ever get there.

Public rush for personal genomic data

So what can we expect to see immediately prior to a public rush for personal genomic data of ‘sufficient’ quality? One perceptual change that’s required is the value of polygenic analyses over monogenic. The ‘direct to consumer’ DNA testing market, and in particular companies that customise your exercise or nutritional regimen, have been promoting their ‘one SNP-one result’ methodology for years. In reality, we know that genetic variants tend to work in concert to produce a spectrum of phenotypic outcomes, but it’s always been easier to sell results along the lines of ‘if you have this mutation, your caffeine metabolism will be faster than average’. Fortunately, companies like DNAFit and Fitness Genes are pressing ahead of the rest by publishing papers showing a polygenic approach is more effective - they argue that the company with the smartest algorithms will produce the most effective results, and they’re right to an extent.

Having taken hundreds of tests, I’ve always been pleased to see that most ‘direct to consumer’ DNA testing companies caveat their results in the face of ‘environmental’ or ‘non-genetic’ factors –laudable given the area’s so under-regulated. Some of the best tests will give a genetic predisposition to a condition and indicate the extent to which environmental vs. genetic factors play a part. In a YouGov survey I ran in 2014, when asking the public how they felt about receiving a genetic predisposition to a condition where their DNA only played a small role, most said they didn’t understand the question. Nevertheless, a stipulation that companies indicate the extent to which their results depend on genetics vs. environment would be welcome - it would help raise the importance of the distinction and allow individuals to derive more meaning from their results. When the In Vitro Diagnostic Medical Device Directive (aka the IVD Directive) is updated in April next year, the MHRA may require companies to do exactly this.

Could epigenetics be a game-changer?

Then there’s epigenetics, a field that threatens to derail the current ‘direct to consumer’ model completely (despite the fact that most DNA tests for physiological traits tend to accurately reflect their customers personal experience!). If the public truly understood that genes could be ‘switched on or off’ then the ‘one SNP-one result’ approach would be widely ignored. But what would raise the awareness of epigenetics? Secondary education aside, I suspect epigenetics will leap into the public consciousness when a ‘direct to consumer’ company starts marketing a ‘switch flipping’ therapy for the first time. At that point, it’d most definitely be in your interest to know the genetic variants you possess, and what the benefits of their expression/repression could be!

Until the general level of understanding rises above these bars, I believe the public desire for personal genomic data will remain niche. From looking at our own DNA Testing Choice site data, the number of consumers researching and buying their whole genome sequence has increased eight-fold in the last year, but these individuals make up just 1% of our total audience, with most researching and buying tests to trace their ancestry, or to give them a customised diet and exercise regimen.

Data sharing incentives

However, there may be a shortcut… There’s a move towards paying individuals for their DNA to fuel genetic research. DNA Simple gives you cold hard cash in return for your genomic data and Genos has recently announced that they’ll do the same… This is a departure from the popular approach to acquiring genetic data from the public, which is for companies to sell a report in return for access (e.g. 23andMe). Another less popular approach is to provide a free report in return for the data (e.g. DNA Land), but this model relies on customers already having a copy of their digitised DNA, which restricts them to the comparatively small pool of consumers who’ve already taken a test. I suspect that when the art of buying genomic data from the public is honed, everyone will take notice. We’ll be looking to these companies to interpret our data as well as pay us for it, but more importantly, we’ll want them to produce the highest quality (and likely epigenetic) data-sets so that we can sell ourselves for even more.

In such a world awash with data of all types and from disparate sources, it’ll be companies like Repositive that allow researchers to take advantage of it. ‘Direct to consumer’ DNA testing companies will be tempted to silo the data they acquire at first, but sharing in return for access to other silos will ultimately be more profitable. Repositive are in the right place at the right time, by registering data to the platform, users can increase data visibility and help link people to it. Helping the community share data and find collaborators is one of the original founding principles that Repositive was built on. I look forward to seeing them wrestle with such a challenge!

By Craig Macpherson - Editor and founder of DNA Testing Choice

==If you are interested in sequencing yourself, whether it is to discover your ancestry, gain insights into your health, or paternity testing; visit DNA Testing Choice and find the service which best suits you.==

Related Blog Posts

The Value of Personal Data

Personal Genomics Open Access Datasets Even More European-Biased Than Scientific Literature?

More Methylation data to power your research