DNA Dilemma: The FAQs

I'm trying to choose whether or not I want to take a direct-to-consumer genetic scan of hundreds of thousands of variants in my genome. Before we get started with the big questions, here are some basic queries that many consumers may have, as well as some information about myself and this project.

So how will this work, anyway?

I'm currently in possession of a direct-to-consumer genetic test—I just haven't decided if I'll use it. For three days this week, I'll pose a question a day to a variety of sources about the value of these tests. I'll post their answers on this site, along with my reaction. I'll also be soliciting opinions from commenters and people following me on Twitter. On Friday, Aug. 6, I'll evaluate everything I've learned and reveal whether I've decided to take the test.

Which DTC genetic testing kit did you buy?

Currently, there are two large, reputable companies offering scans of hundreds of thousands of genetic markers directly to consumers, the same ones that introduced the tests to the public, launching within a day of each other in 2007: 23andMe and deCODE. I don't plan to reveal here which company's kit I bought, because I don't want to become a de facto ad for either company if I take the test. However, it's worth noting that there are some differences between the firms. 23andMe charges a lot less, for instance—$499 for a health and ancestry scan, compared to the $2,000 price of the deCODEme complete scan—and tests for about half as many genetic variants, 550,000 compared to deCODE's 1 million. Aside from the occasional embarrassing slip-up, 23andMe does a fine job of quality control in identifying genes. So does deCODE, which isn't a testing company so much as a lab that happens to sell tests while producing a near-unrivalled body of genetics research. Both companies have also put a lot of effort into conveying their findings in innovative (if not always fully transparent) ways. They try to interpret their data with a lengthy report and continuous updates on the Web; deCODE's service even links to the original research that underlies the test.

How do the tests work?

The company sends you either a test tube to spit into or a "Buccal DNA Collector," which is essentially a cheek swab. You use the kit and send it back; the company puts your DNA on a chip used to identify the four different nucleotides that encode genetic information; and within a few weeks, the company mails you a detailed report about mutations in your DNA and what it thinks they may mean for your health. Bear in mind, though, the company isn't looking at every nucleotide in your genome. It's examining only a few hundred thousand to a million "SNPs," or single-nucleotide polymorphisms, which vary from person to person. Scientists have linked these SNPs to diseases (both common and rare) and physical traits (the most famous "fun" one, assessed by 23andMe, is your likelihood of having wet or dry earwax). In some cases, the company may give you the raw data from its assessment of your genome—a list of every mutation it has identified—which means you can go look up scientific papers about each of the SNPs on your own, if you have the time and inclination.

Are there any DTC tests that offer a more comprehensive look at the genome?

A few companies, such as Knome and Illumina, offer services that scan and interpret a lot more variants than a few hundred thousand—they look at the whole genome, not just a selection of SNPs—but their prices start at $19,550, and I make a reporter's salary. The cost of sequencing is falling fast, but it still has to go a lot lower before these services are within the reach of the average person—and by then regulation may be so heavy that the companies won't sell their services directly to consumers.

Are all DTC tests made by reputable firms?

As you'd expect in an unregulated market, there are companies making highly questionable claims about science they either don't understand or are willfully mischaracterizing. Compared to 2007, the variety of genetic tests now publicly available is mind-boggling. While few companies sell the kind of genome-wide scan I'm interested in, many others—thousands, even—offer narrower gene-scanning services. There are Web sites that match you with romantic partners and tests that twist legitimate and fascinating new papers into a ludicrous "estimate [of] how much Neanderthal is in your ancestry." You can even get a cheek-swab test that "provides scientific confirmation of the physical characteristics, behavioral tendencies, personality traits, and potential health risks" of your dog. (From the SkyMall catalog, no less.) And, since the FDA hasn't been weeding the DTC garden, there's always a niche, as Steven Pinker puts it, for "companies [who] terrify hypochondriacs by turning dubious probabilities into Genes of Doom."

A recent GAO report and Congressional hearing showed that different companies sometimes come to different conclusions from the same DNA. Doesn't that mean the companies are untrustworthy?

Personal genomics is a new phenomenon with rough edges. The risk-based predictions differ because the companies sometimes base their numbers on different estimates of risk in the population, and there's no standard list of genetic variants worth analyzing, so they test for different genes. The companies and the FDA are both reportedly keen to come up with a standard list of variants (they've tried before, but discussions fell apart), and recently, they've been responding to criticism, saying they'll explain to consumers more clearly how they arrive at their conclusions.

Still, even once standards are developed, risk prediction will remain an imprecise practice. Scientists are very far from knowing all the genetic factors that influence health, to say nothing of all the environmental ones. This point, too, got somewhat misconstrued at the hearing. For instance, the GAO report features a customer who was told he had a low risk of irregular heartbeat based on the genes identified in his test—despite the fact that, unknown to the testing company, he'd had a pacemaker implanted for that condition 13 years earlier. This sounds scandalous and was presented as such. It's not. It's merely a sign that the science is incomplete and that the guy had either bad luck or bad health habits. After all, genes are not destiny.

How did you pick the experts whose opinions you solicited?

Many of them were obvious choices. For instance, people who follow this field know that for well-informed analysis of genomics and the law, the go-to guy is Dan Vorhaus, who we'll hear from on Thursday. But I want to bring in fresh voices, too, which is why I'm hoping for widespread participation from readers in the comments section and on Twitter (#DNAdilemma).

If you've already bought the test, doesn't that give you an incentive to go on and take it so you don't waste your money?

It could. Similarly, getting the test results for free—i.e., getting NEWSWEEK to reimburse them—would warp my ability to judge whether I found them worth the cost. So here's what I'm going to do: If I decide to get my DNA analyzed, I won't ask for reimbursement. I'll just pay for the service myself. If I decide not to take the test, NEWSWEEK will pick up the tab, I won't lose any money, and it'll be like I never bought the kit in the first place.

What's that book you're reading in the picture?

It's a 14-year-old edition of An Introduction to Genetic Analysis, a common medical school textbook. I have it because my husband bought it as a med student. You can get the current version here. I wouldn't recommend the old edition unless you're the sort of person who likes to comb through the indexes of outdated texts, laughing at how little people once knew (there's not even an entry for the Human Genome Project!). Apparently I am this sort of person.

Follow Mary on Twitter as she makes her decision, or add your opinion using the #DNADilemma hashtag.

DNA Dilemma: The Complete Series

Monday: Is Now the Time to Test?
Tuesday: What Can I Learn From This Test?
Wednesday: How Meaningful Are The Results?
Thursday: Should These Tests Be On the Market?
Friday: My Genome, Myself: I Make My Decision.
Recommended Reading: Further Articles, Essays, and Websites
DNA Dilemma FAQs: You're Doing What, Exactly?