Isolation and identification of age-related DNA methylation markers for forensic age-prediction

Abstract

Age-prediction is an important part of forensic science. There is no available method of individual age-prediction for general forensic biological samples at crime scenes. Accumulating evidence indicates that aging resembles a developmentally regulated process tightly controlled by specific age-associated methylation exists in human genome. This study isolated and identified eight gene fragments in which the degree of cytosine methylation is significantly correlated with age in blood of 40 donors. Furthermore, we validated two CpG sites of each gene fragment and replicated our results in a general population sample of 40 males and 25 females with a wide age-range (11-72 years). The methylation of these fragments is linear with age over a range of six decades (Fragment P1 (r=-0.64), P2 (r=-0.58), P3 (r=-0.79), R1 (r=0.82), R2 (r=0.63), R3 (r=0.59), R4 (r=0.63) and R5 (r=0.62)). Using average methylation of two CpG sites from each fragment, we built a regression model that explained 95% of the variance in age and is able to predict the age of an individual with great accuracy (R(2)=0.918). The predicted values are highly correlated with the observed age in the sample (r=0.91). This study implicates that DNA methylation will be an available biological marker of age-prediction. Furthermore, measurement of relevant sites in the genome could be a tool in routine forensic screening to predict age of biological samples.

Keywords: Age; Correlation, Prediction; DNA methylation; Forensic marker.