Development of the VISAGE enhanced tool and statistical models for epigenetic age estimation in blood, buccal cells and bones

DNA methylation is known as a biomarker for age with applications in forensics. Here we describe the VISAGE (VISible Attributes through GEnomics) Consortium's enhanced tool for epigenetic age estimation in somatic tissues. The tool is based on eight DNA methylation markers (44 CpGs), bisulfite multiplex PCR followed by sequencing on the MiSeq FGx platform, and three statistical prediction models for blood, buccal cells and bones. The model for blood is based on six CpGs from <i>ELOVL2</i>, <i>MIR29B2CHG</i>, <i>KLF14</i>, <i>FHL2</i>, <i>TRIM59</i> and <i>PDE4C</i>, and predicts age with a mean absolute error (MAE) of 3.2 years, while the model for buccal cells includes five CpGs from <i>PDE4C</i>, <i>MIR29B2CHG</i>, <i>ELOVL2</i>, <i>KLF14</i> and <i>EDARADD</i> and predicts age with MAE of 3.7 years, and the model for bones has six CpGs from <i>ELOVL2, KLF14, PDE4C</i> and <i>ASPA</i> and predicts age with MAE of 3.4 years. The VISAGE enhanced tool for age estimation in somatic tissues enables reliable collection of DNA methylation data from small amounts of DNA using a sensitive multiplex MPS assay that provides accurate estimation of age in blood, buccal swabs, and bones using the statistical model tailored to each tissue.