Sometimes the genetic results shown in DNAPassport or BABYGlimpse (for example, your predicted eye color) may not exactly match what you'd expect. Fortunately, there's no need to worry about the integrity of your sample because Helix maintains the most rigorous standards for sample handling in their state-of-the-art CLIA- and CAP-certified DNA sequencing laboratory. So why could your results look a little different than you expected? Well, it comes back to genetics and a little math!
Some background info: Using genotypes to predict phenotype
DNAPassport and BABYGlimpse are DNA-powered apps that provide insight into a variety of ancestry, appearance and wellness traits. To deliver your insights, the app uses a person's genotypes (the As, Ts, Cs and Gs in your DNA) to estimate a phenotype (the observable characteristic or trait, such as brown hair).
For each of the traits shown in DNAPassport and BABYGlimpse, a prediction is made based on a variety of genetic markers. Some of these traits, such as eye color, combine the results of several genetic markers into a “score” that ultimately produces the results seen in the app. DNAPassport uses statistical models from scientific publications (cited in the References section in the app) to calculate that score based on a users’ genotype at these different genetic markers. We use the most up-to-date scientific research available for predicting a trait based on genetics, but research is constantly ongoing to improve these predictions. For more information, check out our other article about trait prediction here.
So why might the app show a different result than I expected?
There may be some cases where the genetic markers measured are unable to explain the traits we display in real life. In this case, the statistical model may produce an unexpected result. One interesting example of this can be seen with eye color: green eye color tends to be more difficult to predict using genetic markers than brown or blue eye color so the app may display slightly different percentages for some customers with green or hazel eyes. We are actively working to improve these and other predictions in DNAPassport and BABYGlimpse, so you may see an improved result in the coming months.
It is important to keep in mind that, for all traits in DNAPassport and BABYGlimpse, there may be other factors which are not accounted for in the statistical model, such as additional genetic factors beyond the few the product measures, or environmental and other factors. If you click on "Learn more" text next to any trait and scroll down to the bottom, you can view the Genes vs Environment graph that summarizes this information. In this graph you can see the heritability of the trait (e.g. "90% genes") which represents the proportion of the variability in a trait in a population that is attributable to all genetic variation. On the left side of the graph, the dark green box labeled "Scientifically Understood" represents the degree to which the specific markers in DNAPassport or BABYGlimpse are able to explain phenotypic variance for a given trait. For an in-depth explanation of heritability and phenotypic variance, check out this article.
For some traits in DNAPassport and BABYGlimpse, such as cilantro taste, the result is based on a user’s genotype at a single genetic marker which has been associated with that trait. For these traits, you'll notice that the dark green "Scientifically Understood" box is very small (e.g. 3%). This is because there are many potential reasons why someone could enjoy or not enjoy cilantro that aren’t related to this one particular genetic marker, and therefore aren’t measured by DNAPassport or BABYGlimpse. These factors could be genetic, or could be due to non-genetic environmental factors, such as nutrition, allergy, or many others. So, while the research supporting an association between a genetic marker and a trait is valid, the overall impact of genetics on how that trait is expressed may still be very small.
Still have questions?
HumanCode is always striving to deliver the best product experience and most accurate predictions available. If you have additional questions regarding your DNAPassport or BABYGlimpse results, chat with us via the chat function (lower right of this screen) or email us at firstname.lastname@example.org.