Provision of a ‘biological witness’ to a crime is currently the focus of much forensic genetics research involving information on visible characteristics and ancestry, and provision of a reliable prediction of age of the donor of a crime stain is a useful addition to this. The research exploits epigenetic changes, such as additions or loss of methyl groups to cytosine nucleotides that are associated with age. The extent of methylation is quantified by converting non-methylated cytosines to uracil through bisulphite treatment and their subsequent conversion to thymine when amplified, assessing the ratio of the two nucleotides in target areas. Because forensic material is limited the research must identify and model a small number that are highly correlated with age yet not affected by differences in tissue origin, sex, origin and disease to provide a prediction of chronological age with minimal error.

To date this research has focused on accurate age prediction of all ages from trace blood spots. Current methods predict an average error of 3.3 years, reducing to 2.6 years in people under 55 years of age. There is an increasing interest in being able to accurately determine whether an individual is an adult or a child. Current assessment of age relies on holistic assessment to indicate whether or not the physical appearance and demeanour of an individual strongly suggests they are significantly over 18, but applicability and potential bias when applied to a group of asylum seekers has led to doubts about the reliability of this approach, particularly where presenting individuals are not European. While measurements of bones and teeth through the use of X-rays can be useful, our future research will focus on developing markers from a non-invasive source, such as buccal swabs, and limited to a range of around ten to twenty-five years of age.