An Orchestra of Genes
Imagine your genome like an orchestra. Sometimes, sound flows from just one instrument, a gifted soloist that single-handedly fills the concert hall. But more often, the music is an amalgamation of many different instruments, each one making a small contribution to the overall tune.
The genome’s roughly 20,000 genes work in a similar way. A particular trait or disease can be the work of a single gene — for example, blood type and cystic fibrosis. But the most common conditions and traits, such as heart disease, diabetes, height, and intelligence, are the result of multiple genes functioning in combination. In fact, hundreds, even thousands of genes can work together, each one exerting a slight influence on human biology. Today scientists are developing ways to better dissect these polygenic traits — that is, which genes and genetic variants contribute and how much?
A common approach is to calculate polygenic scores, which are the sum of (usually small) effect sizes of many genetic variants. Polygenic scores can help determine an individual’s genetic propensity toward a particular trait or disease. In some cases, these polygenic risk scores are becoming useful tools for doctors— for example, identifying patients who are likely to suffer a heart attack despite having low cholesterol. However, it is important to remember that genes tell only part of the story. The vast majority of health conditions are also influenced by environmental factors including lifestyle.
The New Nebula Library
At Nebula Genomics, we are committed to bringing cutting-edge scientific discoveries to our users, as well as making their meaning more intuitive to grasp. Previously, the Nebula Research Library contained studies with the discovered genetic variants listed in tables. However, because everyone typically carries a different combination of many genetic variants, it can be difficult to interpret what a list of genetic variants mean in the context of one’s own genome.
To address this issue, we have updated the format of the Nebula Research Library. Now, in addition to providing tables that list discovered genetic variants, we calculate two numbers for most of the studies. The first number is a polygenic score. It summarizes the effect of all genetic variants discovered in the study and present in your genome. The second number is a percentile that is calculated for the polygenic score. It tells you how your polygenic score compares to the scores of other Nebula Genomics users
Figure 1 shows a typical Nebula Research Library entry. This user has a polygenic score for attention deficit hyperactivity disorder (ADHD) that was calculated from ADHD-associated variants identified in this study (A). This score places the user into the 97th percentile (B), which means that he has a higher polygenic score for ADHD (based on variants identified in this study) than 97% of our users.
The first two columns of the table did not change in this update (C). They contain the variant IDs and the alleles – versions of a variant – that have been found to be associated with a trait or disease, which is ADHD in the example in Figure 1.
Next, come three new columns that we added in this update (D). The first column lists effect sizes which are the contributions of different alleles to the polygenic score. Some alleles have positive effect sizes and increase the score (highlighted in green) while other alleles have negative effect sizes and decrease the score (highlighted in blue). Alleles not present in your genome do not affect your polygenic score (not highlighted).
The allele frequencies in the second column tell you the percentage of people in the population that carry each of the listed alleles. From low-coverage whole-genome sequencing or imported genotyping data, we can reliably infer most alleles that are relatively common (frequency of > 5%). Results for rare alleles are not shown and these alleles are not included in the calculation of the polygenic scores.
The third new column shows the statistical significance for each of the associations discovered in a study. These numbers are called p-values. The smaller the p-value, the more certain it is that the discovered association between a trait and a genetic variant is real. We include only variants that show a highly significant association with a trait and sort them in the tables from most to least significant.
The last column in the table did not change in this update. It shows your genotype – the two alleles you have inherited from your parents – and an estimated accuracy of our prediction. Note that we account for this uncertainty when calculating your polygenic scores. If the accuracy is low, then the contribution to the polygenic score is reduced.
Proceed with Caution
We would like to emphasize that the scores that we calculate are much less predictive than polygenic risk scores that are used for diagnostic purposes. The latter are carefully calibrated and often incorporate millions of genetic variants — covering nearly the entire swath of the genome — whereas our scores include only a small set of the most significant variants. This means our polygenic scores are a rough estimate and should only be used for educational purposes. They are not intended for medical or diagnostic use. In addition, as we described above, genes are not necessarily destiny. If you have a particular variant associated with a disease in your genome, it does not mean you will develop that disease. The same goes for polygenic scores. A high polygenic score / percentile for a particular disease does not necessarily mean a significantly increased disease risk, because many genetic effects remain undiscovered and environmental factors almost always play a major role. As always, if you have any questions about your health, please seek the advice and input of a healthcare provider.
Join Us Today
Enter the age of personal genomics and learn about how the latest discoveries relate to your DNA by ordering our low-coverage whole-genome sequencing or get started for free by uploading your 23andMe/AncestryDNA data!