DNA甲基化
肋软骨
软骨
逐步回归
医学
生物
计算生物学
计算机科学
遗传学
内科学
解剖
基因
基因表达
作者
Ju Yeon Jung,Moon Hyun So,Kyu‐Sik Jeong,Sang‐In Kim,Eun‐jin Kim,Ji Hwan Park,Eung‐Soo Kim,Hwan Young Lee
标识
DOI:10.1111/1556-4029.15470
摘要
The DNA intelligence tool, DNA methylation-based age prediction, can help identify disaster victims and suspects in criminal investigations. In this study, we developed a costal cartilage-based age prediction tool that uses massive parallel sequencing (MPS) of age-associated DNA methylation markers. Costal cartilage samples were obtained from 85 deceased Koreans, aged between 26 and 89 years. An MPS library was prepared using two rounds of multiplex polymerase chain reaction of nine genes (TMEM51, MIR29B2CHG, EDARADD, FHL2, TRIM59, ELOVL2, KLF14, ASPA, and PDE4C). The DNA methylation status of 45 CpG sites was determined and used to train an age prediction model via stepwise regression analysis. Nine CpGs in MIR29B2CHG, FHL2, TRIM59, ELOVL2, KLF14, and ASPA were selected for regression model construction. A leave-one-out cross-validation analysis revealed the high performance of the age prediction model, with a mean absolute error (MAE) and root mean square error of 4.97 and 6.43 years, respectively. Additionally, our model showed good performance with a MAE of 6.06 years in the analysis of data of 181 costal cartilage samples collected from Europeans. Our model effectively estimates the age of deceased individuals using costal cartilage samples; therefore, it can be a valuable forensic tool for disaster victim and missing person investigation.
科研通智能强力驱动
Strongly Powered by AbleSci AI