A novel approach for age estimation based on blood mitochondrial DNA analysis.

Age estimation is a critical area of research in forensic medicine. Accurate age assessments provide vital insights for criminal investigations and play a significant role in shaping case characteristics, as well as influencing conviction and sentencing outcomes. In previous studies, skeletal and dental morphology was the most commonly employed method for age estimation. However, these techniques have notable limitations, including insufficient accuracy, significant subjectivity, and challenges in application to trace evidence. Consequently, the search for new molecular markers for age inference has emerged as a prominent research focus. Recently, mitochondrial DNA (mtDNA) has attracted attention due to its close association with development and aging processes.However, mtDNA has not been applied in forensic age estimation. In this study, the relative quantity of mitochondrial DNA to nuclear DNA, i.e., the mitochondrial DNA copy number(mtDNAcn), was determined using real-time quantitative PCR and used as an indicator for assessing mitochondrial copy number. The correlation between mtDNA and age was further investigated. A model for estimating age from mtDNAcn in blood was constructed through machine learning techniques. Of the various methods evaluated, the k-nearest neighbors (kNN) algorithm demonstrated the best performance, achieving a mean absolute error (MAE) of 4.938 years and an R² value of 0.758. This framework offers a novel approach for age estimation from trace forensic evidence and significantly expands the potential applications of mtDNA in the field of forensic science.