Calculating age with DNA
Blog posted by: Hira Javaid, DPhil student
How would you react if I told you I could calculate your age if I took a sample of your blood?
I wouldn’t be lying!
Every day our cells are exposed to damaging agents that can cause harmful changes which accumulate over time and lead to ageing related diseases such as cancer. Quantifying these changes has allowed scientists to develop models that calculate a person’s biological age, a measure of the “well-being” of their cells.
The Horvath clock is one of the most widely-used age clocks and uses DNA methylation data to calculate age (Horvath S, Genome Biology, 2013). DNA methylation is a type of epigenetic change that refers to the addition of specific groups called methyl groups onto DNA at specific sites. The methylation state of the DNA contributes to whether certain genes are switched on or off, similar to post-it notes in a recipe book directing you to the recipes you’d like to open in the near future! Each methylation site can either be methylated or unmethylated and thus can be considered as a number – 1 for methylated and 0 for unmethylated. The Horvath clock makes use of this and uses the pattern of methylation across someone’s DNA to extract the pattern of 1s and 0s. Across the whole of someone’s DNA, the methylation pattern can be written as a code of sequence of 1s and 0s and fed into a computer. The computer can then compare the actual sequence with patterns from the very start of life and patterns from people at various stages of life. An algorithm can then try to calculate the likely biological age of a person by fitting the changes observed in their case into the pattern we see from other people.
The Horvath clock works surprisingly well, on average calculating age to within 3 years of a person’s real age. It is not entirely understood what the clock measures, as it does not relate to a person’s biological age but to chronological age. Usually cells keep track of age through the number of times they divide, however, the Horvath clock also works for non-dividing cells such as neurons!
One might wonder what the use of age clocks may be if a person can simply be asked for their age. One use is in forensics, e.g. to predict age at crime scenes. Studying the methylation sites involved in ageing also allows us to understand the ageing process and develop and test anti-ageing interventions. More recently, DNA methylation has even been used to predict time to death! (Horvath et al Aging, 2016) In terms of application to cancer, an important aspect of methylation age is the discrepancy between a person’s methylation age and real age. The greater this difference, i.e. the “age acceleration” of a person, the greater their risk of developing ageing-related diseases and this includes cancer. Thus, this measure represents a potentially important implication to cancer risk stratification and earlier diagnoses.
Age clocks are an example of what can be achieved when experts from different fields work together. In this case exploring the methylation pattern requires both biologists, statisticians, and computer scientists to work together. The future is all about such interdisciplinary work.