DNA甲基化分析血液及血痕推断个体年龄的法医学探索研究

Age prediction of an individual based on biological traces remained in a crime scene is of ultimate importance for the criminal investigation. The increasing evidences indicate some CpG sites reveal the DNA methylation is almost linearly changed with aging, and thus may serve as an accurate tool for age prediction. As age-related DNAm changes might be a highly tissue-specific, blood and bloodstain are the most common biological traces remained in crime scenes. It is necessary to establish a methylation assay which may yield a robust age prediction. The bisulfite pyrosequencing technology was proved as an accurate and stable approach in methylation quantitative analysis in our preliminarily study. However,it is still necessary to further improve our DNA methylation assay by recruiting new AR-CpG markers and utilizing a robust new model. In this study, genome-wide DNA methylation profiles from younger donors (aged from 10 to 25 years) and from senior donors (aged from 55 to 65 years) were analyzed to identify AR-CpG sites. 100 unrelated Chinese Han volunteers aging between 10 and 75 were used to validate these age-associated CpG sites by bisulfite pyrosequencing approach. Then establish the age prediction model by using different models including multivariate linear regression, support vector recognition (SVR) and relevance vector recognition (RVR). Based on the mean absolute deviation (MAD) from real chronological age of different models, we can determine the age prediction model as well as our DNA methylation assay. Validation test was also carried out to determine whether our DNA methylation assay is suitable for age prediction in bloodstains. Ultimately, provide the scientific basis for its application in forensic criminal investigation

表观遗传学的DNA甲基化研究中发现一些AR-CpG位点与年龄存在线性关系，可以应用于年龄推断。AR-CpG位点存在组织特异性，血液和血痕是犯罪现场最常见的生物性检材，因此构建一个基于血液DNA的甲基化年龄推断体系是非常必要的。在此前的基础研究中我们构建了基于“重亚硫酸盐转化+焦磷酸测序”和“多元线性回归”的甲基化年龄推断体系，为进一步提高体系的推断效能，我们需要纳入新的AR-CpG位点并对模型算法进行改进。本课题拟通过甲基化芯片筛选适用于中国汉族人群的AR-CpG位点；利用“焦磷酸测序”技术建立甲基化数据库并验证AR-CpG位点；分别采用多元线性回归、支持向量机回归及相关向量机回归建模，比较各模型MAD值大小，确定DNA甲基化年龄推断模型；验证该模型在血痕DNA样本中的适用性。以期最终构建一个更加稳定、准确的适用于血液和血痕的DNA甲基化年龄推断体系，为其在法医学领域的应用提供科学依据。

根据犯罪现场中残留的生物物证进行犯罪嫌疑人的年龄推断对于案件调查至关重要。表观遗传学的DNA甲基化研究中发现一些AR-CpG位点与年龄存在线性关系，可以应用于年龄推断。AR-CpG位点存在组织特异性，血液和血痕是犯罪现场最常见的生物性检材。前期我们构建了基于“重亚硫酸盐转化+焦磷酸测序”和“多元线性回归”的甲基化年龄推断体系。本课题以DNA甲基化中的与年龄相关AR-CpG位点为基础，开展纳入新的AR-CpG位点并对模型算法进行改进的法医学年龄推断探索研究。通过甲基化芯片筛选、寻找适合作为中国汉族人群年龄推断的AR-CpG位点，共找到4个AR-CpG位点作为年龄推断的新标记，利用“焦磷酸测序”技术建立甲基化数据库并验证AR-CpG位点；根据采用多元线性回归、支持向量机回归及相关向量机回归建模后确定DNA甲基化年龄推断模型；验证该模型在血痕DNA样本中也可适用。最终构建了更加稳定、准确的适用于血液和血痕的DNA甲基化年龄推断体系，为其在法医学领域的应用提供科学依据。