Our Algorithms

DNA methylation analysis has significantly expanded in the field of longevity since the publication of the Horvath pan-tissue clocks in 2013. However, it is important to note that not all testing and analysis is created equal.

There are some very important considerations to look at when vetting a company for Epigenetic Analysis.  

1. What type of algorithms are used?
2. What type of tissues are collected?
3. How much data is generated from my testing?
4. Can I access my raw data for independent analysis?
5. Will there be updates to my data as new insights are published?

We Deliver More than Just Biological Age

We also provide a full suite of aging related metrics. This includes telomere length measurements, intrinsic and extrinsic age calculations, immune cell subset deconvolution, and the current pace of aging. 

We also try and give you more than just data, with a full epidemiologic review of studies published on age-related methylation. With these studies included in your TruAge reports, we hope to give context to what may have influenced your various aging scores. These studies can also act as a jumping-off point for investigating possible lifestyle changes to improve your aging.  

Our Priorities:

Large Scale Data Generation
We measure 900,000 CpGg locations on the genome to provide one of the most robust testing services available. We base this analysis on the EPIC850k methylation array by illumina which is routinely used in the majority of methylation informed clinical trials.
Robust QA/QC Analysis
Our Lab and Bioinformatics teams adopt rigorous procedures to make sure our reports are accurate. We utilize consistent and ongoing quality improvement methods to improve everything from sample collection and benchtop protocol, to automation and data analysis.
Correct tissue type collection:

Every type of tissue has a different epigenetic signature. As a result, the type of tissue you test is very important, and can influence the data you get from analysing it. For instance, if we were to test the tissue from your cerebellum, we would get much younger ages than if we tested blood. We also would get much higher ages if we tested breast tissue.

Almost all of the published algorithms for longevity based epigenetic clocks used whole blood, when they trained the algorithms for biological aging. As a result we prioritize blood collection as the primary type of analysis to make sure your results are as accurate as possible.

Licensing and creating the most relevant aging algorithms:
We are proud to work with academia to license the best algorithms in the field of aging. Below, you will see the algorithms which we use and their importance for understanding a complete picture of your aging process.
Providing patient access to the raw data:
We also believe that you should completely own and have access to your own data. As a result, we are able to give you your methylation data in 2 ways. We provide a report which lists out the CpGs used in most published aging algorithms. We also provide access to your raw data file (.idat) upon request so that you can own and control your own data for future analysis.
Continual Updates:
We understand that new developments are being made every day in the field of epigenetics and aging. We constantly work to improve the insights available in our testing. You can expect regular updates that add to our reports at no additional charge (unless we have to license the algorithm).
Referrals to medical practitioners to help you interpret:
We also know that understanding how your epigenetic aging data and its impact to your life can often be difficult. As a result, we often refer our users to medical providers at no charge. We do this so that they can help you make sense of these metrics and what it means to you.

Our Exclusive Algorithms

DunedinPoAM4 Current Rate of Aging

The DunedinPoAm algorithm is a one of a kind algorithm created by researchers from Duke, Columbia, and the University of Otago.

Duke professors Terrie Moffitt and Avshalom Caspi head a team of six who developed the DunedinPoAm tool this year. Building the data base took the international team five decades; while they tracked biological changes in the bodies of 1037 amazing New Zealanders who are members of the Dunedin Multidisciplinary Health and Development study, a project that began with their birth in 1972. “We are now applying DunedinPoAm in 19 other large health-tracking studies. One goal is to test just how sensitively it detects when people change their lifestyles and health behaviours. We are looking at many thousands of people: different ethnic groups, age groups, and men and women, living in different countries. DunedinPoAm is the only aging measure so far that was trained on biological change, and the enthusiasm from the international teams who are participating is super exciting!” said Moffitt.

It is a great tool to add to your longevity based analysis because, unlike biological age clock algorithms, it is able to tell you how you are aging at the precise moment of the test versus the overall age of your body. This helps you know if you are currently implementing the best protocol to reduce your biological aging and disease risk over time.

Telomere Length Estimator

Telomere attrition is one of the 9 hallmarks of aging. It was previously the main molecular marker for the aging process. 

However, it’s correlation value to age has always been relatively low compared to epigenetic markers of age. Additionally, telomere length has not always been very predictive of the outcome of disease.

Dr. Horvath’s telomere algorithm on telomere length prediction showed double the correlation to age that telomere length alone could provide. It also showed that DNA methylation measurements of telomere length we more predictive of many types of conditions such as death, coronary heart disease, and congestive heart failure. 

Our telomere length prediction algorithm shows similar trends and accuracy of prediction 

Telomere Length Diagram - showing a cell nucleus, chromosome and telomere end-caps

Epigenetic Biological Age

We have one of the largest private epigenetic databases in the world, with over 13,000 patients tested to date.

By collecting important covariates, we’ve created a state of the art biological age predictor. Building on science already in the literature, we have used our large database to create a highly accurate biological age model with a high correlation value to age (r=.98). 

Our Epigenetic Biological Age Report was the first report we offered to the public, and expanded to what we offer today. 

Intrinsic and Extrinsic Age

Intrinsic and Extrinsic Epigenetic Age are ways to look closer at Epigenetic Biological Aging, to discover the impact that different systems on your body have on aging – especially your Immune System and Proteome

These algorithms take immune cell ratios into account when examining whole blood. They work together to give an advanced look into your epigenetic aging and how your immune system, cellular makeup, and hormonal biology is changing your epigenetic age. 

Completed TruDiagnostic Algorithms Being Submitted for Publication

Advanced T-Cell Immune Cell Subset Deconvolution

Skin Deconvolution Method

Saliva Deconvolution Method

Death Prediction Algorithm

Multi Omic Analyte Prediction Algorithms

Vetting Epigenetic Testing Companies

What type of algorithms are used?

The most common way to measure DNA methylation is calculating beta values, which are generated using Illumina’s beadchip array and quantifying them from a series of mathematical equations.

The data generated from your methylation analysis results is something called a beta Value. The definition of the beta value is defined below and essentially provides us an estimate of the percentage of methylation at a certain location on the genome.

The beta value is the ratio of the methylated probe intensity. The beta values are a score between 0 and 1 and can be interpreted as the approximation of the percentage of methylation for the population of a given CpG site in the sample.

The beta values are then subsequently used to construct the log-likelihood function of a generalized regression model that relates beta values to linear functions of the covariates.

What type of tissues are collected?

Unfortunately, epigenetics can get very complicated when we start discussing multiple types of tissues. This is because while every cell in our body has the same DNA, each cell has many different epigenetic markers. 

This means that each cell type we test will give us very different epigenetic methylation signatures, even when they’re all taken from the same person on the same day.

These signatures are measured by something called a beta value which is a relative intensity of methylation.  We measure 900,000 beta values for each of our tests. 

However, algorithms are used to interpret these beta values into actionable data.  That is how we take beta values and objectively convert them to things like telomere length or biologic age.  

To date, all published algorithms have only been validated from blood.

As a result, that is the only tissue which we will use for our analysis.  If we were to use these algorithms in breast tissue instead, we might get vastly older ages.  Brain tissue usually would be the opposite and give us much lower ages.  Therefore, for accurate results, blood is the best type of tissue to test!

If you are testing in other tissues we would recommend looking at validation datasets!

How much data is generated?

The more methylation values that are tested, the more data is able to be analyzed for insight into your epigenome.  This is important as most researchers use the same epigenetic infrastructure to run many of their clinical trials.  As a result, any new development can be interpreted from the data we have generated from your samples in the past. This allows us to keep you updated with the best scientific information as it relates to you!

Additionally, testing more CpG locations on the DNA allows us to create better and more accurate algorithms.  This will continue to improve as investigations into epigenetics become more cost effective and more expansive over time. This will allow us to use even more data to create the most predictive algorithms for age and beyond. 

Can I access my raw data for independent analysis?

Your data is your property. This means you should have complete access to it. Having access to your beta values for these specific CpG sites gives you the freedom to use your epigenetic information however you would like. You can even use your beta values in published algorithms! This gives you control over how you want to interpret your methylation values beyond what we report out.

We are committed to total transparency in how we handle and report your data. We also want to deliver your data to you in digestible formats.

While we believe that the interpretation of these health metrics is important, we also believe that you should know your methylation status. 

That’s why we provide you the raw data we use in this analysis.

Will there be updates to my data as new insights are published?

The beauty of storing your data in our repository is that we are able to run future reports based on your methylation values from the sample you have already provided. These future reports are created as soon as relevant epigenetic research has been published, which adds prodigious value to our test at no additional cost to you.