(UPDATED: See below.)
Personal-genomics vendors like 23andMe and deCODEme, which promise to give ordinary individuals a peek at their genetic inheritance, have received a ton of press attention since they launched last November (not least of all from us — see here and here for starters). Unless you happened to have a spare $1,000 laying around, however, you were pretty much out of luck if you simply wanted to know exactly what you might be getting for your cash.
Until now, that is. On Monday, deCODEme — the precious capitalization is intensely annoying, but I’ve bowed to the inevitable, since even the NYT uses it now — began offering a “demo user” that allows anyone to poke around in the results from a reference genome. While it obviously doesn’t tell you anything about your own particular genetic makeup, it can certainly help give you a sense of what sort of information the service provides for your $985 (which, to be clear, is an “introductory price” — deCODEme hasn’t said what it will charge once the “introduction” is over).
My capsule summary: At this stage, deCODEme doesn’t appear to offer a heckuva lot of value for what it’s charging. The service will presumably grow and evolve over time, but for now, you’re probably best off exploring other options — or simply waiting for the technology to improve and prices to drop.
AI Weekly
The must-read newsletter for AI and Big Data industry written by Khari Johnson, Kyle Wiggers, and Seth Colaner.
Included with VentureBeat Insider and VentureBeat VIP memberships.
Like 23andMe, deCODEme essentially offers its customers a kind of shorthand scan of their genomes — not a full reading of the six billion DNA “letters” (technically, base pairs) that make up our 23 pairs of chromosomes, but a gene-chip analysis that identifies roughly one million sites where the letters are known to vary between individuals. In a rough sense, each such variation — technically known as single-nucleotide polymorphisms, or SNPs — determines which version of a particular gene you have.
That information, in turn, may shed light on your susceptibility to particular diseases or predispositions toward certain psychological traits or physical attributes. (A cautionary note: The science of linking gene variation to actual individual characteristics like this is still in its infancy, and many experts — including some who recently held forth in the New England Journal of Medicine, caution against reading too much into these relationships. deCODEme also advises against making medical decisions based on its information — a good thing, too, as we’ll see.) To get the analysis, customers swab the inside of their cheek with a high-tech Q-tip, then mail it in to deCODEme, which collects the cells from the swab, runs the analysis, and posts the resulting information in a supposedly secure location on the Web.
Activating the reference genome (it can also be reached via a link on the deCODEme home page, although it’s fairly inconspicuous) brings you to a page — click on the thumbnail at left for a larger view — with four basic tools. “My Gene Profile” lets you browse through a variety of disease conditions — 18 at the moment — and to calculate your known genetic risk of developing them. Other tools predict your hair and eye color based on your genetic makeup, trace your ancestry and measure your genetic similarity to 53 ethnic groups from around the world, and let you share and compare your genetic profile, or genotype, with any “friends” or family members who’ve also joined up with deCODEme.
At first glance, that looks like a lot of information, but it doesn’t take long to find the deCODEme system intensely frustrating. (For the sake of readability, I’ll describe the demo genome as if it actually represented my own results — which, of course, it doesn’t.) The gene-profile section, for instance (see the thumbnail at left), lists the 18 conditions for which deCODEme thinks there is enough hard evidence linking genes to disease to calculate your genetic risks. Clicking on any of the diseases, a list that runs from age-related macular degeneration, or AMD (vision loss in the elderly), to type 2 diabetes, takes you to a description of the disease and several tabs that outline your particular susceptibility to the condition, the scientific underpinnings for the risk calculation, and a list of possible risk factors and preventative steps you might take in order to ward off your “genetic fate.”
All nifty enough. A quick look through the first several conditions reveals that “I” have an especially low risk of ending up with Alzheimer’s disease or celiac disease (an autoimmune gastrointestinal disorder), but a higher-than-average chance of going blind or developing asthma. (See “my” Alzheimer’s risk thumbnail to the left.) The interface, however, is incredibly clumsy — checking your risk factors for each of the 18 conditions here will take a while, as the site requires you to click into each one, then click again for your risk calculation, and then click out again to get back to the main gene-profile page. I gave up after looking through the first six, although I’d undoubtedly be more curious were this my actual genotype.
Worse, at least for information gluttons like me, the “scientific details” available for each condition are pretty thin, and are presented in a way that isn’t likely to make sense to anyone but geneticists. (See thumbnail at left.) Many of the risk calculations seem to be based on a single SNP, which means they are bound to be misleading, since the evidence to date strongly suggests that most diseases are the result of interplay between dozens or hundreds of genes, any of which might either elevate your risk or offer protection against the disease. And there doesn’t seem to be any way to search for particular SNPs in your genotype, which would be useful if you happened to see a recent scientific disease-risk finding and wanted to see how it might affect you. (UPDATE: I’ve taken a closer look at all of deCODEme’s disease-risk calculations here.)
One final nitpick before moving on — the gene-profile page offers two additional viewing options, both oddly implemented and, at least in their current form, basically useless. For instance, it’s possible to identify diseases that affect particular organs by clicking on a human-body graphic, although for some reason Alzheimer’s disease isn’t listed when you select the head. (In fact, it’s not included in any of the highlighted organ systems.) You can also choose to view disease traits that are associated with one of your 23 chromosomes, a feature for which I still can’t figure out a use unless you happen to be a biology student.
The “physical attributes” tool is, at the moment, little more than a joke (thumbnail at left), as it does nothing but “predict” your hair and eye color. (Unsurprisingly, given that the reference genome is most likely Icelandic — deCODEme’s parent company is based there — “my” hair is likely red and “my” eyes are probably blue.) Presumably this section will offer more information as scientific knowledge accumulates, but at the moment it’s not clear why deCODEme included it at all.
Meanwhile, the “ancestry” page (left) serves up some very hard-to-decipher graphics that are supposed to illustrate your genetic commonality with the rest of humanity, but which just ended up confusing the heck out of me. (I subsequently figured out that the large numbers on each pop-up balloon rank your degree of relatedness with the various geographic regions, so “I” am most closely related to people of European descent, then those from Southwest Asia, then East Asia, and so on.) Clicking a balloon gives you further ethnic-group detail from that particular part of the world. For the moment, this and the related “ancestry” items look like curiosities and little more.
Finally, there’s “Compare Me,” the supposed heart of the social network deCODEme hopes to build around people and their genotypes (thumbnail at left). To say this function looks lame would be one heck of an understatement. Compare yourself to a friend or relative from the list and you get a cool-looking but almost information-free chromosomal map that purports to illustrate the genetic regions you have in common with the other individual, only without providing any information beyond colored bars on blobby chromosomes. (You can zoom in, but that just makes the colored bars bigger and the chromosomes blobbier.) A “relationship check” link promises to determine how closely related you might be, but anyone less related to you than a parent or child yields a major rat’s nest of possible family ties — for instance, click here to see what it returned when I compared “myself” to “my” nephew.
Disappointingly, one of the potentially interesting features here should let you download your genotype, but unfortunately it’s not enabled for the demo account.
Now, it’s entirely possible that deCODEme may not have implemented its most up-to-date features on the demo account, although I can’t see why it wouldn’t — the idea is presumably to entice users to sign up for the service, not to show them a crippled version and still expect them to cough up some serious bucks. It’s equally possible that the service could radically improve in short order over what the company is showing here.
If so, I’m more than willing to take another look. For now, though, I have to say that if I’d just plunked down $985 for deCODEme, I’d be royally pissed, both at the waste of money and at the lack of information, flexibility and user-friendly functionality here. I’m kind of astonished that deCODEme thinks this version of its service resembles a finished product in any way, shape or form.
In any case, I’ll review 23andMe if they ever follow through on earlier indications that they might also provide a reference genome for exploration, as well as any other related services such as Navigenics or Knome. (By the way, the Houston Chronicle recently reported that two other Texas companies, Family Tree DNA and Seqwright, also plan to launch personal-genomics services. Hat tip to the Genetic Genealogist.)
Have you signed up for deCODEme? Share your experience in comments.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn More