Genetic testing: hacking your intimate secrets

Genetic testing companies and the reliability of the genetic information have come under scrutiny, again. But the accessibility of these tests bring privacy implications, especially when such testing actually becomes accurate and provides meaningful info?  Will we see genome hacking?

Direct-to-consumer genetic testing companies have come under fire again, this time from some of their own, rather than the usual suspects: governments and regulators.

Scrutiny is good and these recommendations and regulations may be OK for the amount of knowledge we can extract from our DNA today (not much, relatively).  But  in the next 20 or so years it is highly probable we will be able to reveal some intimate details about ourselves from a few cells left lying around.

Direct-to-consumer genetic testing companies are those that will take your DNA (usually saliva or cheek swab) and a heap of your $$ and reveal a bunch of your genetic secrets: your disease risk, IQ and even the ability to taste bitterness. Kits can be purchased online, you cough up your sample (literally) and post it off.

Craig Venter and colleagues published a critical article in Nature highlighting many of the recognised issues with direct-to-consumer testing (along with doing a bit of their own comparative testing).  They also present a number of recommendations to help alleviate these issues, which appear sound, though I am sure there will be some disagreement as people analyse them more closely.

One of the issues identified is that some of the genetic links to traits or diseases are tenuous at best (eg, IQ). And there is minimal follow up with the recipient about what to do with the often ambiguous test results.

Nature’s editorial response to the paper says it best: “This means that customers will frequently receive results telling them only that they face the ambiguous possibility of a somewhat elevated risk of a little-understood disorder.”

And just because a test says you have a low risk of a particular disease doesn’t necessarily mean you can kick back and relax.  Traits or diseases can be controlled by many genes, not just the handful of marker genes we have discovered (Marker genes are genetic sequences that are associated with a particular trait – not necessarily a gene, but they tend to be associated with one that is linked to a trait). And for those we do know about, we often have a limited understanding about how big an influence they have on a trait. For example, studies have revealed more than 40 genetic regions involved in height. Added together, these variants account for just 5% of our variation in height – see Nature article.

Venter and co’s paper explore some of the technical aspect including the manipulation and different use of statistics by different testing companies.  This gets tricky to explain and I am no expert so I refer you to the article.

Another point they raise is that different companies tend to use different genetic markers giving a different risk profile for each person.  This is because as explained above, each trait can have many different genes, variations of the same gene or genetic markers associated with it. The consequence is, as Venter found out, that you can find you have a high risk of a disease with one company’s test and an average or low risk with another company.  One minute you are panicking and wrapping yourself in cotton wool, the next breathing a sigh of relief and cracking open the chocolate bar– or left scratching your head.

What they did find, was that where the two testing companies used the same markers, their disease predictions were comparable.

Their investigation resulted in nine recommendations, which include the following three:

A focus on high risk diseases

That is, those markers that can accurately place you in a high risk of acquiring a disease. Customers could make lifestyle changes based on these.

Agree on strong-effect markers

Direct-to-consumer companies need to agree on using a core set of strong-effect markers to achieve better consensus and consistent reporting to us recipients.

Monitor behavioural outcomes.

Aimed at communities rather than the companies, they suggest funding for more public studies to monitor behaviour resulting from this testing to identify effective strategies that use personal genomic data to improve a person’s health.

Finally, forget spooks and spies, what about when we do get a better knowledge of our genes and how they are regulated? We will have the capability to be a lot more definitive about a lot of diseases and personal traits, traits we ourselves might be uncomfortable knowing about, let alone strangers with potentially nefarious motives.  With testing easily accessible even today, what will that mean for privacy, or ourselves personally. That is, how will we handle such knowledge? Can we handle the truth?.  And it is dead easy to steal someone’s DNA and uncover another’s intimate genetic secrets– see Genome hacking blog

For Australians today, I am unsure what having a test means as far as disclosure goes to entities such as insurance companies. As far as I am aware, an insurance company cannot force a person to have a genetic test, but if you do have one you are required to disclose that information if you want to be covered.  I will do some homework.

Similar topics are discussed on other GNTIS blogs

Jason Major

GNTIS