DNA App for Couples Knows My Kids Hair Color
Reproduction is important in science. One experiment may be a fluke, but if you get the same results over and over, you know you are getting closer to the truth. So when a DNA app called BabyGlimpse claimed it could predict what my kids would look like, I was ready . I have three children, so I can check their results in triplicate.
BabyGlimpse is one of the billion-dollar DNA testing services launched last year in partnership with Helix . You send a saliva sample to Helix, and then Helix sequences your DNA and stores your data. From there, you can allow partner companies to view any subsets of your data relevant to that product. Insitome will tell you how connected you are with the Neanderthals. Vinome can check if you have a sweet preference marker and recommend a wine you might like. BabyGlimpse looks for several DNA markers from you and your partner to guess which baby you might have together.
At first glance, it sounds dumbly romantic, like the modern equivalent that casts your name along with the name of a loved one. But it costs $ 260 (or $ 99 if each of you has already sent a saliva sample to Helix) and includes showing some of your DNA data to your partner. Just on a fairly superficial level – there are no genes for Alzheimer’s or anything else in the results – but it’s always helpful to think about how much information you really want to share.
The usual caveats regarding DNA testing also apply. BabyGlimpse looks at your ethnicity, so you might see data that goes against family lore and raises uncomfortable questions. And all of this only works if you don’t give a damn about Helix, so you need to be confident about Helix’s privacy policy.
How it works
Helix appears to be holding back its partners’ claims pretty tightly . So you won’t find overt pseudoscience, but you may well find things based on weak or partial evidence from the scientific literature.
While Helix sequences large portions of your genome, analysis from partners like BabyGlimpse works in the same way as genotyping services like 23 and I. They scan your DNA for individual letters that scientific studies have found to correlate with certain traits. For example, with two Europeans cytosines (CC) at a place calledrs4988235, usually nottolerate lactose. But there is uncertainty here: without this variation, it is possible to have lactose intolerance, and this marker has no predictive value if your ancestors were not European.
So even for simple signs like lactose intolerance, you should be suspicious of your results. BabyGlimpse advises that my husband and I are probably both lactose tolerant and can drink as much milk as we like; our kids will probably be too. (They eat well milk, but they are still too young to know for sure. Even babies with lactose intolerance usually do not show symptoms at first.)
But no one sends their saliva on a romantic evening to see if their kids will be lactose intolerant. The company’s screenshots highlight the color of the hair and eyes, as well as the thought of catching a glimpse , right? your unborn child.
What i have learned
Before my husband’s results came in, I could view my own. The app suggested that I have a 68% chance of having brown hair. That’s right – I was blonde as a kid, but now I have brown hair. However, this was not true with my eyes: his prediction was blue, probably only 27 percent of my true color, something from the walnut green family.
When my husband’s results came in, they said he had brown hair and brown eyes (yes), and according to the app, our kids probably do too. This is a victory: all three of my children fit this description.
There is no single DNA marker for hair or eye color, as many genes are involved in making brown pigment and transporting it to the right places. Thus, BabyGlimpse calculates points based on several different DNA markers.
But the best part is that they put uncertainty in context. Some traits are purely genetic; it is almost always possible that something in your diet, parenting, prenatal experience, or something we haven’t discovered yet plays a role. There can also be an element of pure luck. When it comes to hair color, BabyGlimpse says 86 percent of this trait is due to genetics, and the markers they use could explain 70 to 80 percent of that. The appendix says that eye color is 96 percent genetic, with markers covering 80 to 90 percent of that color. One of the articles they cited says that this level of prediction is possible for people from all over the world, making it more useful as a predictor than a lactose intolerance marker that only says anything about Europeans.
My husband tans quite well and has slightly darker skin than mine, but our results show that he has “very light” skin, while my children and I have “fair” skin. I don’t know what the difference should be between them, but okay.
The height did not seem to be different. BabyGlimpse said I am above average and my husband below. Based on this calculator, we are both above average, bigger than me. In any case, our children must be tall, and this is more or less confirmed.
The prediction of hair curl was incorrect, but information was also lacking. Instead of breaking down the probabilities of straight, wavy, and curly hair, he simply said that all of us and our children will have “wavy” hair. This is true for me and one of the children. My husband and one of my sons have very curly hair, and the third child has hair that is too short to be distinguished. BabyGlimpse admits that known DNA variations only cover “a relatively small proportion (~ 20-50%)” of the variability in curly hair. So this is not a particularly useful test.
There is a prediction of male pattern baldness, which is not exactly the trait we look for in babies. To be honest, all these traits are related to the appearance of an adult. Many newborns have blonde hair, blue or gray eyes, and they are very, very short. So it doesn’t really give you a mental picture of your baby, but rather gives you more of what to expect 20 or more years from now, long after you forget your BabyGlimpse results.
Things get a little weirder when you fall into the “just for fun” category. Based onone negligible marker that has been tested on professional athletes, BabyGlimpse describes our children’s “muscle fiber type” as “normal muscle composition”. You will have to work hard (like most of us). ” Call me a skeptic (who, am I?) But I don’t think this is a very useful result. Especially when they acknowledge – or perhaps guess – that this marker explains less than 5 percent of the genetics involved.
According to BabyGlimpse’s interpretation of just one marker, our babies may also be slightly addicted to sweets. They will also have the personality of either a “strategist” or a “mixed warrior-strategist,” according to another tiny variation . According to SNPedia, which I use to link to these markers, 23andme describes the dichotomy as “alarming” and “warrior”. Anyway, I really don’t think that my child’s personality is determined by one nucleotide.
BabyGlimpse, to their credit, retains its individuality in the “just for fun” section of the app and points out that the marker they tested explains only “about 1%” of known genetic variation, which is only 60% of the total variation. But then, if you know that the result explains only 0.6 percent of the characteristic you are talking about, what is the point?
What we got from this
BabyGlimpse is right in promoting itself just for fun; I cannot think of a single meaningful decision that I would make based on these results, and this is probably for the best. But what would we do without it? We probably thought that our children would have brown hair, brown eyes and fair skin, and still did not know if they would turn out to be “warriors”, “strategists” or elite athletes.
So we didn’t know anything. But was that fun? And now we can blame the children’s DNA (which really means blaming ourselves) when they share our sweet tooth. On the other hand, we did it anyway.