Frequently Asked Questions

This site is a tool for anyone interested in better understanding polygenic scores. In practice, the two use cases for what we've built are individuals interested in learning more about their genetic predispositions, and parents interested in learning more about their embryos. It was for the latter group that this site was originally built. There are an increasing number of companies, such as Orchid Health and Genomic Prediction, which offer parents the ability to screen their embryos for disease risks. Unfortunately these companies do not offer predictors for other traits that contribute to quality of life, such as intelligence or educational attainment. This site aims to fix that by providing open source predictors that any parents or individuals can use.

Navigate to the Upload tab, click the 'Upload' button, and select your genomic data file. We support various file formats including VCF and 23andMe raw data.

We currently support VCF files and raw data files from major genomic testing services like 23andMe, AncestryDNA, and MyHeritage.

You can request a copy of your embryo's genomes from the genetic testing company. Once you receive this data, you can upload it to this site to learn more about each embryo. You can then make a choice about which embryo to implant based on both the genetic predictors provided by the genetic testing company and the supplementary predictors provided by this site.

Simply put, their predictors are better than ours. They are trained with more data and validated more thoroughly. However, their services are quite expensive. Our aim is not to provide a superior product, but to offer something that's easily and cheaply available to any parent or individual that wants to use it. We're still working on quantifying the exact size of the difference in our predictor and those offered by paid services, but my best guess is the expected IQ gain from using our predictor is roughly 50-70% as large as that offered by the one paid service we're aware of.

Your results go through three stages. First, scientists study the DNA of hundreds of thousands of people and look for genetic variants that tend to show up more often in people with a certain trait. They figure out how much each variant nudges the trait up or down. This is done independently of our services and before you upload any genetic data. Second, we look at which of those variants you carry and add up all the nudges to get your raw polygenic score. Third, we compare your score to a reference population and tell you percentage of people you score above. If you're at the 75th percentile, your genetic predisposition is higher than about 3 out of 4 people in that reference group.

Heritability is how much of a trait is driven by genetics versus everything else. For height, genetics explains about 80% of why some people are taller than others. For something like depression, genetics plays a smaller role (around 30-40%), with life experiences mattering more. This matters because heritability sets the ceiling for how good any genetic predictor can be. If a trait is only 30% genetic, even a perfect predictor could only explain 30% of the picture. Traits that are both highly heritable and well-studied (like height) have the most reliable predictions.

When comparing multiple genomes, you can tell us which traits matter most to you using a 0–3 scale. Setting a predictor to 3 means 'this is very important to me,' while 1 means 'this matters less.' Setting it to 0 removes that trait from the comparison entirely. The overall score takes each genome's percentile for every predictor, factors in your weights, and averages them into a single number for easy ranking. Keep in mind: this score reflects what you said matters most—someone else with different priorities would get a different ranking from the same data.

The score represents how far above or below average a genome is across all selected traits. A positive score means better than average overall, zero means average, and a negative score means below average overall. Disease traits like cancer and diabetes are automatically inverted so that lower disease risk contributes positively. The score is centered around the 50th percentile (average), so even a 'good' genome can have a negative score if most of its traits fall below the 50th percentile.

Your genomic data is encrypted using industry-standard encryption protocols before being stored. Only you can access your data through your secure account.

Yes, you can delete your data at any time through your profile settings. Once deleted, all your data is permanently removed from our servers.