How accurate your results are, and the predictions for you baby, depend on a wide range of factors.   We do our best to explain them below.

For the actual DNA testing:

We have partnered with Helix for DNA sequencing.  This portion of the process takes your saliva sample as an input, and outputs your genetic sequence data.  Helix uses state-of-the-art “next gen sequencing” technology from Illumina, the leader in DNA sequencing.  HumanCode then uses this data to produce the insights you get from our products.  You can learn more about Helix, the sequencing technology, and their CAP accredited and CLIA certified lab at  The CAP and CLIA certifications are both standards designed to ensure quality laboratory testing.

How accurate is sequencing?

Next-generation sequencing is typically 99.999% accurate at each location of the DNA. Helix has built a state of the art sequencing lab, and they also work closely with Illumina, the market leader in next generation sequencing technologies. Their lab is optimized to provide the highest level of accuracy as they sequence your DNA. Even with all this work, there are some limitations to the technology. In the rare case that Helix is unable to read small portions of your DNA, they will mark that section as "unreadable" rather than guessing. You should feel confident that when you look at your data, it is as accurate as possible.

Learn More =>

How “accurate” a given insight (about you, your partner, or baby prediction) may be depends on a variety of factors including but not limited to:

  • Genes vs environment:  each trait may be influenced to some extent, in you, your partner, and any of your offspring by both genetics and environment.  The genetic portion of this is called “heritability” (what we get from our parents).  Some traits have high heritability and low environmental influence.  Ancestry is an example - your environment plays no role in your ancestry.  Height is another example:  height is known to be roughly 70% heritable, meaning your genes determine 70% of your height, and your environment (e.g. nutrition) determines the other 30%.
  • Strength of research: Within the heritable component, the current research may explain only some portion of the variability of a person.  Of the 70% of height that is known to be heritable, the research we used to make the prediction explains roughly 20% of this.  Part of this limitation is due to limitations in how much research that scientists have done, to date.   As more genetic research is done, the science should explain more and more of the heritable element of each trait.  HumanCode updates our prediction algorithms based on new science.
  • Reference population: in some cases, the research we used for the prediction algorithm may be based on a reference population which is different from yours (e.g. research done on people with European ancestry, but you are of African descent).  Again this will improve as more research becomes available.
  • Penetrance: what percentage of people with given genetic markers, actually have the condition that has been associated, is known as penetrance.  Very few genetic markers are “100% penetrant”, meaning it’s common for someone to have a genetic marker linked to a higher probability of having a certain trait, without having it.
  • Expressivity: even for genetic markers that are 100% penetrant, the size of the effect of the trait or condition within the individual may vary.  For example, you could have a genetic marker for a certain allergy that is 100% penetrant - everyone with the genetic marker has allergy symptoms.  But within an individual, the severity of the allergy could vary widely.  
  • Baby Simulation: predictions for Baby are subject to all of the above, plus additional uncertainty.  For most traits, we predict the Baby’s genetics by running thousands of simulations of how the parent genetics could combine.  Based on the outcome, we select the combinations that appear most frequently as the highest probability.  There is a large amount of unknown variability in how the genetics combine, in some cases it's simply random.
  • Specifics of Genotypes and Traits: the combination of the parents’ genetics, and how the actual trait manifests, can also produce a higher degree of certainty or uncertainty.  For example, if there is a trait that is tied to a single genetic marker (or SNP), and the trait outcome is based on a homozygous allele (meaning, having AA, or TT at that position, but not AT), and both parents carry the same homozygous allele (both are AA), then we know the baby will "for sure" inherit AA at that location in the genome.  But even then there are exceptions...
  • Variants: genetic variants which can include “mutations” that occur during the process of the DNA of parents combining, are known as “de novo” mutations.  These are not uncommon, and if they happen at a certain genetic location, could produce an outcome that is completely new to the family genetic lineage.  A de novo mutation could cause a child to have red hair, for example, even if neither parent carried the genes for this trait.

In general, the trait predictions are simply our “best guess” based on the available research.

We do our best to explain how accurate the prediction is, within the Trait Detail experience in the product.

A big part of the BABYGlimpse experience is learning more about how genetics and inheritance of traits work, and we're focused on delivering that promise regardless of how much of the science a person understands.

Did this answer your question?