Six-CpG saliva epigenetic clock tested in 76 Han Chinese samples
Using saliva methylation data from 76 Han Chinese individuals, researchers selected six CpG sites and reported age-prediction errors of about 3.24 and 3.27 years for two models, comparable to leading clocks in validation. [1]
Automatic summary · not editorially reviewed
The key findings
The study used genome-wide 850K methylation array data from saliva samples of 76 Han Chinese individuals. [1]
Robust Lasso and Support Vector Regression had mean absolute errors of 3.240 and 3.267 years. [1]
Model errors were lower than those reported for 85% of 60 previously published saliva-based epigenetic clocks. [1]
The source at a glance
- Study type / format
- Abstract of a model-development and validation study.
- Population / context
- Saliva samples from 76 Han Chinese individuals.
- Summary based on
- Study abstract
- Original publication
- June 23, 2026
Saliva samples from 76 Han Chinese donors
The study aimed to build a precise epigenetic clock that estimates a saliva donor's age, motivated by forensic use of saliva. It used genome-wide 850K methylation array data from saliva samples of 76 Han Chinese individuals. From these data and a literature review, the authors first selected 33 candidate CpG sites. This is an abstract; the full text was not reviewed.
Supporting source [1]
Six CpG sites chosen by machine learning
A multi-stage selection process ran 250 iterations across five machine learning algorithms in parallel: Elastic Net, Lasso, Ridge, Support Vector Regression and robust Lasso. Recursive feature elimination aimed to reduce mean absolute error and identified six core CpG sites strongly associated with age. Among the models, robust Lasso and Support Vector Regression gave the best and comparable accuracy, with mean absolute errors of 3.240 and 3.267 years, significantly better than the other linear models.
Supporting source [1]
Validation against three datasets and seven clocks
Validation on three independent datasets and comparison with seven established epigenetic clocks showed accuracy comparable to leading methods such as the Hannum and Horvath clocks, while substantially outperforming others such as the Levine, PedBE and Wu clocks. The models' mean absolute errors were lower than those reported for 85% of 60 previously published saliva-based epigenetic clocks. The authors describe the six-site panel as valuable for forensic age prediction and aging-related research; these are reported associations and predictions, not proven causes or lifespan benefits.
Supporting source [1]
How to interpret this
Abstract only; the full text was not reviewed. The sample was 76 Han Chinese individuals, and validation used three independent datasets. Reported accuracy is predictive error, not proof of causation or of any health or lifespan benefit.
Based on the study abstract; the full paper, tables and supplementary data were not reviewed.
Source & supporting passages
Teng W, Chu X, Sun W, Huang A, Wang Z, Zhu B, Liu X. · europepmc.org · June 23, 2026