Age estimation from DNA methylation in sectioned pulp-free teeth

Single, fragmented, and pulp-free teeth are challenging specimens for forensic age estimation, particularly when conventional morphological indicators are unavailable or compromised. DNA methylation provides a molecular approach for age prediction and may be useful for degraded dental samples. In this study, we established and evaluated methylation-based age-prediction models using sectioned pulp-free teeth as simulated forensic specimens. In this study, we prioritized 10 previously reported age-associated CpG loci that have been validated in skeletal tissues, dental tissues, and ultimately constructed methylation-based age-prediction models using nine CpG sites significantly correlated with chronological age in incised and pulp-free teeth (n = 128). Before model construction, the effects of tooth types and sex were assessed. Neither tooth types nor sex significantly modified the methylation-age relationship according to ANOVA-based interaction analysis, with P values of 0.340 and 0.468, respectively. Adding methylation-by-tooth types or methylation-by-sex interaction terms also did not significantly reduce prediction error, with P values of 0.210 and 0.156, respectively. Age estimation models were developed using partial least squares regression, ridge regression, elastic net regression, and support vector regression with radial basis function kerne (SVR-RBF), with hyperparameters optimized via ten-fold cross-validation to prevent overfitting and ensure stability and generalizability. Ridge regression provided the best balance between fitting and generalization, with test-set R² = 0.749 and MAE = 6.112 years. Paired analysis of pulp-retained and pulp-free teeth showed no significant difference in overall prediction error, but cg22507023 differed significantly between the two groups. Supplementary validation on paired teeth of distinct positions from 12 randomly selected individuals confirmed no intra-individual differences in CpG methylation level and predicted age. Overall, this study demonstrates that DNA methylation analysis can provide useful molecular evidence for forensic age estimation in pulp-free teeth and supports unified modeling without stratification by tooth types or sex.