“Our study demonstrates a method for predicting a person’s ‘internal’ time of day that could eventually be used to help diagnose and monitor circadian and sleep-related disorders, and also to personalise treatments,” said lead author Jake Hughey from Vanderbilt University in Tennessee, US.
In addition, ZeitZeiger can also be used to show how disruptions to sleep-wake and light-dark cycles — by modern environments, for example due to shift work or reduced exposure to sunlight — affect the circadian clock.
Circadian dysfunction is linked to conditions such as cancer, depressive disorder and obesity.
Developing treatments that improve the function of or that account for the circadian system has the potential to improve multiple areas of human health, the researchers stated in the paper published in the journal Genome Medicine.
For the study, the team used 498 publicly available samples from 60 individuals who had blood drawn either throughout the day on a normal sleep-wake and light-dark cycle (controls), or following disruptions to their sleep-wake and light-dark cycle (condition).
Using their machine learning algorithm to analyse the RNA in the blood samples the researchers were able to identify a set 15 of genes capable of accurately predicting a person’s circadian time.
Out of these 13 genes were not ‘core’ genes — necessary for the generation and regulation of circadian rhythms — of the human circadian clock.
The results showed that ZeitZeiger could achieve state-of-the-art accuracy and also detected when the circadian clock was phase-shifted or dysfunctional.