The Naked Science of a DNA test

Nowadays, a lot of companies offer online ancestry tests, or tests to quantify your risk of inheriting some life-changing diseases. But how seriously should we take their results? Izzie Clarke and Tom Crawford spoke to Garrett Hellenthal from UCL and Julianna Cintron from 23andMe in order to find out…

Tom - Fill the tube with saliva to the black wavy line… shall we just crack on?

Izzie - Okay! You get the idea. But in order for companies to analyse our genetic information, all we have to do is spit into a test tube and send it off to the lab. Tom sent his to a company called 23&Me to find out about his health and physical traits, and I wanted to explore my family history. So, how does a bit of my saliva reveal so much about my ancestry?

Garrett Hellenthal from University College London’s Genetics Institute…

Garrett - Well, it’s because that saliva contains your genetic code so it contains a series of cells, and each of those cells contains your complete genetic information. So what these companies do is they extract that genetic information from the cells in your saliva, and then they take that DNA and they identify a series of genetic markers that define your unique DNa sequence. They look at about 500,000 or so different pieces of genetic code, and then they’re going to take your genetic code and compare it that of the people that they have in their database and they identify who do you share lots of matching DNA patterns with. If you share lots of matching patterns with particular people from a particular area, that suggests that you may have ancestors that come from that area or, more accurately, it reflects the fact that you have descendants that now live in that area.

Izzie - That’s how it then corresponds to a place, for example?

Garrett - Exactly, yes. We don’t necessarily know that your ancestors came from that area themselves, but it could be true that they at least have descendants that live in that area. So you have to kind of make an assumption that if you’re matching to people in those areas, then the ancestors came from those areas as well.

Izzie - I’ve got my results and I’m slightly confused by some of them. I am 30% Irish and I guess that’s not too much of a surprise considering that a few of my great-grandparents were actually Irish.

Garrett - No, I don’t think so. With your four grandparents you’re expected to get about 25% of your DNA from each one of them, so your if you look back at your great-grandparents you get half of that, so you get about 12.5% from each one of them. If you have a couple of great-grandparents that came from Ireland you would expect to get about 25% of your DNA inherited from them.

Izzie - Okay, thats good. Then we’ve got 24% Western European and 15% Great Britain, so that's pretty accurate I would say, and then I’ve got 15% Scandinavian and I’m not entirely sure where this comes from. Does this mean someone in my family has been a bit naughty?

Garrett - It could mean that, but it doesn’t have to mean that. Because if we think of our ancestors back in time, we know that our great-grandparents, or great-great-grandparents; they have several children who might have other children and so on and so forth, but that this person has some descendants that live in the UK and some that live in Scandinavia, but it also could reflect that migrations from Norway into the UK. In particular during the 9th and 10th centuries there were a series of Norse/Viking migrations from Norway, for example. There’s also been anglo saxon migrations in the 5th and 6th centuries from areas around North Germany. All of these would be considered Vikings in some form and they did contribute and intermix with the local population.

Izzie - I’m just going to assume that I’m part Viking. But what about health and physical appearance? Let’s take a look at Tom’s results…

Tom - I think most of them are correct for me like I should likely have lighter eyes, and I have blue eyes. It says likely little upper back hair, and I can fortunately report I have minimal upper back hair. I was also pleased to see that I’m likely not to have a bald spot; I really hope that one’s true.

Izzie - In addition to appearance, Tom’s test was able to look at specific parts of his genetic code and explain the likelihood of there being a change called a ‘variant’, which could possibly lead to a life-changing illness.

Julianna - My name is Julianna Cintron and I’m a produce specialist on the customer care team at 23andMe.

There are definitely some traits that are more influenced by genetics than others. Having red hair is one of those traits where we found that there is a gene associated with having red hair, and so if you have two copies of that gene you’re more likely to have red hair. It doesn’t mean that you’ll definitely have red hair but it increases your chances.

Tom - Just with you mentioning there this idea of it’s to do with the confidence in something, or there’s a probability. We’re using phrases like more likely, it’s not sort of fixed. In my result I was told that I have a particular variant which leads me to be at a higher risk of Alzheimer's disease and I was just wondering what exactly does this mean? Does this mean I will get Alzheimer’s; does this just mean I’m above average likely; how does this result actually relate to the risk factor?

Julianna - We’re looking at variants that affect your likelihood of developing this disease. It does not mean that you’re going to get Alzheimer’s; we’re not a diagnostic service, so that’s the first thing to keep in mind, we’re looking at the e4 variant. The e4 variant is known to impact your risk of developing Alzheimer’s disease, so having one copy of the e4 variant just puts you at a slightly increased risk of having Alzheimer’s by a certain age. Having two copies of that would increase your risk a little bit more than just having that one copy.

Izzie - We’ve got to be a bit careful about how we look into these results, especially when it comes to health. Back to Garrett…

Garrett - Most of the genetic risks that we’ve seen thus far are relatively small, especially compared to environmental factors like smoking, or unhealthy eating, or just unhealthy lifestyles. So even if you carry a risk variants, you still may not be at much more increased risk relative to people who don’t carry that variance, and it probably very much pales in comparison to how your risk would increase if you’re say a smoker. You don’t want people getting these results and saying: ah, I don’t have the risk factors therefore I can do whatever I want. You also don’t want people getting these scores saying: oh no, I’m at risk of this so I’ve got to completely change my lifestyle. I think there has to be a lot of discussion about how people interpret these results and how best to relay that information.