Forensic use of DNA analysis relies strictly on comparative grounds: DNA profiles obtained from crime scene material are compared with those of known potential suspects. In the future, genetic prediction of externally visible characteristics (EVCs) from crime scene DNA samples is expected to assist police investigations in suspect-less cases, by concentrating on specific groups of people. Among EVCs, age is the most difficult to conceal. A clue to DNA-based age prediction can come from epigenetics, i.e. the study of modifications of gene function or cellular phenotype caused by mechanisms other than changes in the DNA sequence itself. The epigenetic phenomenon of DNA methylation (DNAm) is well known to change during ageing, and various recent studies identified sites which are either hyper- or hypomethylated depending on age.
Since crime stains are usually associated with minimal DNA yields and come from different tissues (e.g. blood, saliva, semen, vaginal mucosa, skin, etc.), the ideal DNAm-based age prediction test must interrogate a limited amount of sites, all with homogeneous methylation patterns across tissues. Our goal is to perform a comprehensive genome-wide DNA methylation analysis, using the Illumina Infinium HumanMethylation 450 BeadChip technology covering >400,000 CpGs in the human genome, on a large training set of donors providing different forensically relevant tissues. By statistical evaluation of methylation values derived from genome-wide array data, a minimal set of age-specific differentially methylated sites (aDMs) providing maximum prediction accuracy across tissues will be identified, and a regression model for age-prediction including mean absolute deviation will be calculated.
Finally, selected aDMs will be combined in a multiplex single base extension (SBE) assay for the detection of methylation status in bisulfate converted DNA, using conventional capillary electrophoresis, and then tested on an independent evaluation set of tissue samples from donors of known age