DNA Methylation Assay Combines Semen Identification With Donor Age Estimation
A new forensic DNA methylation assay combines semen identification and donor age estimation, offering additional information for biological evidence analysis.
Researchers at Sichuan University and collaborating institutions in China have developed a DNA methylation-based assay that can simultaneously help identify semen traces and estimate the age of the person who contributed the biological material.
The study, published online in Forensic Science International on October 29, addresses two questions that can arise when biological evidence is recovered during an investigation: what type of body fluid is present, and what information can it provide about its donor?
The researchers developed a multiplex SNaPshot system incorporating seven age-related CpG sites and two semen-specific markers. DNA methylation refers to chemical modifications at particular DNA positions. These patterns can vary with age and tissue type, making them useful for forensic investigations.
The system was evaluated using 40 semen samples and 19 non-semen samples. The authors reported high specificity for semen identification within the evaluated sample set and found that the assay remained effective with 50 nanograms of DNA input.
For age estimation, the researchers developed five machine-learning models using a historical dataset of 136 semen samples. The models were independently evaluated using 34 semen samples from the study cohort. A linear-kernel support vector machine achieved a mean absolute deviation of 4.648 years among valid age predictions.
The team also tested 20 simulated mixtures containing semen. Twelve were classified as semen detected and eight as semen not detected. Age estimation on evaluable mixtures produced a mean absolute deviation of 8.865 years.
The findings suggest that combining body-fluid identification and age estimation could provide additional investigative information from a single reaction. However, the results remain preliminary.
The evaluation involved limited sample sets, and age estimates showed greater error in mixtures. The assay does not identify a specific individual, and further independent validation across diverse populations and realistic casework conditions is necessary before routine forensic adoption.