Untitled
Apixmed Prism — an innovative next-generation genetic analytics platform
Home keyboard_arrow_right Blog keyboard_arrow_right
From a Single Gene to Thousands of Genetic Variants: The New Scale of Apixmed Prism DNA Tests
Basics of Genetics
Publication date:
7 minutes reading

From a Single Gene to Thousands of Genetic Variants: The New Scale of Apixmed Prism DNA Tests

A woman reviewing the digital Apixmed Prism genetic report on a tablet at her desk — main cover image.

For a long time, one of the central tasks of medical genetics was to determine which change in a gene causes an inherited disease, why it is passed down in a family, and how to detect the corresponding mutation. Today, genetic analysis can be used not only by clinicians but also to answer questions for a wider audience, for example: why some people are more prone to high cholesterol, others to iron deficiency, and why someone struggles to recover even from the lightest workouts despite all the advice?

To find answers to these questions, looking at a single gene is not enough, because a full understanding of how the body’s genetic features work comes from a combination of hundreds and even thousands of variants, each making its own small contribution. That is why, to understand the genetic architecture of traits as a whole, it is important to use large-scale genetic studies (GWAS), on which polygenic risk analysis is based. 

This progress is also reflected in the development of Apixmed Prism, which now covers more than 1000 indicators across 26 health areas. 

So let us look at how genetics reached this level, what can actually be learned from polygenic scores, and where their limits lie.

Two Types of Genetic Architecture: Monogenic and Polygenic

The information that DNA can provide about a given trait depends on its genetic architecture — namely, the number and effect size of the variants that contribute to that trait. The genetic architecture determines both the method of analysis and how categorical a conclusion can reasonably be drawn from the result. 

There are two types of genetic architecture:

  1. Monogenic architecture. A pathogenic variant in a single gene plays the main role in the development of a disease. Even so, manifestations can differ depending on other genetic and non-genetic factors: the same variant in two people can produce different clinical pictures. This is a well-studied area of clinical genetics, important for diagnosing rare inherited conditions.
  2. Polygenic architecture. Complex traits and common diseases have it: the contribution is spread across many genetic variants, and the outcome also depends on lifestyle, age, environment, and other factors.

A telling example is human height. It is obvious that it is largely linked to heredity. Yet this trait is governed not by one gene, and not even by a few dozen.

In a study published in Nature in 2022, researchers analyzed genetic data from more than 5.4 million people and identified 12,111 independent genetic variants associated with differences in height (Yengo et al., Nature, 2022). Each of them has a small effect, but together they account for almost all of the heritability of height that can be estimated from common genetic variants.

This is exactly the principle of polygenic architecture: complex characteristics of the body are shaped by a large number of genetic variants rather than a single decisive change. Similar patterns are studied for lipid metabolism, glucose levels, blood pressure, bone density, and many other characteristics.

For more on how a DNA test differs from a blood test, see the article What a DNA Test Is and How It Differs from a Blood Test.

GWAS and Working with Hundreds of Thousands of Genetic Variants

An important step toward understanding the architecture of complex traits came with genome-wide association studies — GWAS (Genome-Wide Association Studies) (Uffelmann et al., Nat. Rev. Methods Primers, 2021).

Their aim is to identify statistical associations between genetic variants and the traits under study. When researchers study, for example, features of lipid metabolism, they compare the genotypes of hundreds of thousands of participants with their measured cholesterol levels and determine which DNA variants are statistically associated with differences in that level.

Such studies became possible thanks to several changes in the scientific landscape:

  • Larger study sizes. Genetic data from hundreds of thousands and even millions of participants make it possible to detect weak associations that cannot be reliably established in small samples

  • Advances in genotyping technology. Modern platforms determine the genotype at hundreds of thousands of markers simultaneously

  • Improved statistical models. New methods estimate the contribution of individual variants more accurately and account for population heterogeneity in the sample.

Together, these changes have produced a large number of reliably established associations — the material on which polygenic scores are built.

However, a statistical association should not automatically be taken as a causal link. A detected variant is not necessarily causal: it may simply be inherited together with a causal variant located nearby. Alleles at nearby positions in the genome are passed on to offspring together, so a statistical dependence arises between them — this phenomenon is called linkage disequilibrium (LD). Separate methods are used to find probable causal variants, in particular statistical fine-mapping (Schaid et al., Nat. Rev. Genet., 2018).

Apixmed Prism report interface displaying a trait card with a population percentile scale and genetic risk summary.

Polygenic Model: Weights and an Overall Score

GWAS provide a list of genetic variants and a statistical estimate of the association of each with a given trait. A polygenic model uses this material: it selects variants, assigns each a weight reflecting the estimated effect of the corresponding allele, and sums the contributions into a single score (Choi et al., Nat. Protoc., 2020). Some calculation methods additionally account for linkage disequilibrium between variants so as not to count the same signal several times.

When it comes to susceptibility to a disease, such a score is called a polygenic risk score (PRS). For quantitative physiological traits, where the concept of “risk” does not apply, it is more accurate to speak simply of a polygenic score. Below we use the broader term, since it covers both cases.

Percentile: Position in the Reference Population Distribution

A polygenic score is a relative numerical value that is interpreted by comparison with a reference population: the value is compared with the distribution in that group to show where a particular person falls within it.

For clarity, the result is presented in percentiles (that is, as the % of the population that has a lower or higher score). For example, the 85th percentile means that the polygenic score is higher than that of 85% of people in the comparison group.

Two things are easy to misread here:

  1. A percentile shows only the position relative to the reference population. It does not give the probability of falling ill: the 85th percentile does not mean an 85% chance of developing a condition.
  2. The reference population is defined by the methodology of a particular model, and the accuracy of the estimate is affected by the genetic ancestry of the study participants on whom the model is built and by the characteristics of the comparison sample. A significant share of GWAS has been conducted in people of European ancestry, and transferring models to other populations can reduce accuracy (Martin et al., Nat. Genet., 2019). Therefore, the same person may receive a different position in different models.

A polygenic score describes the genetic component of predisposition: it does not confirm the presence of a disease and does not show whether it will develop. A high score does not mean inevitability, and a low one does not rule out risk. Actual health is also influenced by other inherited factors, age, nutrition, physical activity, environment, and medical history.

Systolic Blood Pressure as a Polygenic Trait

Systolic blood pressure is one of the well-studied polygenic traits (Evangelou et al., Nat. Genet., 2018). Researchers have identified numerous genetic variants statistically associated with differences in this measure. Individual variants have a small effect on the likelihood of developing disease, so their combination is used to assess the genetic component. In a polygenic model, each selected variant has a weight, and the contributions are summed using a defined methodology. The resulting score can be compared with the distribution in the reference population.

This example shows that no single variant explains why a particular person’s blood pressure is higher or lower. It is the combination of genetic variants that becomes informative. But under no circumstances does any individual polygenic score remove the need to measure blood pressure with a tonometer, nor does it establish a diagnosis: it describes the genetic component of predisposition. The DNA sequence does not change over the years, whereas blood pressure fluctuates daily.

For more on why it is worth returning to a genetic report, see the article Genetic Data Does Not Age: Why It Is Worth Returning to Your Report.

Characteristics for Which Polygenic Scores Are Used

Polygenic scores are built for traits with a sufficient evidence base. There are many such traits, and they relate to different areas of human health, including:

  • Cardiovascular system: blood pressure regulation, lipid metabolism, other characteristics of the cardiovascular system
  • Metabolism: glucose regulation, fat and carbohydrate metabolism
  • Vitamins and minerals: metabolism of individual nutrients.
  • Exercise physiology: endurance, strength characteristics, recovery.
  • Hormonal and reproductive health: hormonal regulation, functioning of the reproductive system.
  • Sleep and circadian rhythms: sleep pattern, chronotype, regulation of biological rhythms.

This list of systems, and hence of areas in which genetic testing can assess inherited predispositions and risks, is not exhaustive: polygenic models are applied to many other characteristics as well. 

However, the polygenic approach is not universal: some characteristics are assessed from a combination of variants in complex models, others from individual variants with a known biological role, which are interpreted one by one. The choice of method depends on the specific characteristic and on how well it has been studied.

In addition, a genetic test does not establish current vitamin levels, does not detect hormonal disorders, and does not assess actual physical fitness. Laboratory tests and medical examinations are needed for that. The role of a genetic test is different — to explore inherited characteristics that are part of the overall health picture.

Number of Markers and Scope of Interpretation

A modern genetic test can examine hundreds of thousands of DNA markers. But the number of markers does not yet show how much useful information a person will receive.

It is important to distinguish two numbers: the number of genetic markers examined during genotyping, and the number of indicators interpreted separately. The same variant can be linked to several characteristics, while a combination of many variants is used to assess a single complex trait.

Therefore, more than 1000 indicators in an Apixmed Prism report do not mean more than 1000 separate genes or 1000 polygenic scores. This is about the breadth of interpretation: the genetic report covers related characteristics across different areas of health and physiology.

For more on how to compare genetic tests with each other, see the article How to Choose a Genetic Test.

The New Scale of Genetic Tests by Apixmed Prism

The new version of Apixmed Prism covers more than 1000 indicators across 26 thematic areas. Each area is available as a separate test, and the full Ultima genetic panel brings them all together.

By the way they are interpreted, the indicators in Apixmed Prism DNA tests fall into two types:

  • More than 760 composite indicators assessed with polygenic models: for each of them, the contribution to the result comes from a combination of genetic variants, the number of which depends on the specific model and the trait under study.
  • More than 280 individual genetic variants that are interpreted one by one and are not part of the composite indicators.

Behind every DNA test, and hence behind every personal genetic report, there is dedicated work: selecting scientific studies and genetic associations, choosing assessment methods, and preparing interpretive materials. The scale has grown along with the volume of this work: the first versions assessed 200 indicators, then 555, and the current one covers more than 1000.

Each indicator in the report is accompanied by an explanation: the mechanism on which the association is based, recommendations formulated specifically for it, and the sources through which the conclusions can be verified independently.

Premium white unboxed DNA sample collection kit by Apixmed Prism set on a light blue background.

From Genetic Data to a Personalized Understanding of Health

Modern research increasingly shows how complex the genetic architecture of many traits is. Instead of searching for a single decisive change, researchers study thousands of genetic variants, assess their combined effects, and refine interpretation methods.

Further progress is linked to combining genetic information with laboratory results, medical history, and lifestyle data. Such an approach can help assess individual characteristics more precisely and make better-grounded health decisions.

Explore your genetic traits with Apixmed Prism — 1000+ health and physiology indicators from a single saliva sample, for life.

 

Learn more about genetic tests by Apixmed Prism → apixmed.com.ua

 

The results of a genetic test are not a diagnosis or a prediction of future health. Genetic scores reflect individual inherited factors and require interpretation in light of other factors.

 

References

1. Yengo, L. et al. (2022). A saturated map of common genetic variants associated with human height. Nature, 610(7933), 704–712. https://doi.org/10.1038/s41586-022-05275-y

2. Uffelmann, E. et al. (2021). Genome-wide association studies. Nature Reviews Methods Primers, 1(1), 59. https://doi.org/10.1038/s43586-021-00056-9

3. Schaid, D. J., Chen, W., & Larson, N. B. (2018). From genome-wide associations to candidate causal variants by statistical fine-mapping. Nature Reviews Genetics, 19, 491–504. https://doi.org/10.1038/s41576-018-0016-z

4. Choi, S. W., Mak, T. S.-H., & O'Reilly, P. F. (2020). Tutorial: a guide to performing polygenic risk score analyses. Nature Protocols, 15(9), 2759–2772. https://doi.org/10.1038/s41596-020-0353-1

5. Martin, A. R. et al. (2019). Clinical use of current polygenic risk scores may exacerbate health disparities. Nature Genetics, 51(4), 584–591. https://doi.org/10.1038/s41588-019-0379-x

6. Evangelou, E. et al. (2018). Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits. Nature Genetics, 50(10), 1412–1425. https://doi.org/10.1038/s41588-018-0205-x

Any questions left?

Leave your contact details — our specialists will contact you shortly. We will help you understand the offers, select a test according to your request, and answer any questions about the product and process.