Abstract
目的
利用MassARRAY和焦磷酸测序两种方法检测的DNA甲基化数据在年龄推断中的差异,探讨两种检测方法的年龄推断计算方法。
方法
应用MassARRAY和焦磷酸测序,分别对65份和62份外周血样本的9个CpG位点的甲基化水平进行测定,运用多元线性回归模型预测年龄,对比预测年龄与实际年龄之间的差异;对比Z-score转化前后两种方法之间的年龄预测差异。
结果
MassARRAY法,65样本集数据Z-score转化前,平均绝对误差(MAD)=2.49岁,Z-score转化后,MAD=2.44岁;62样本集数据Z-score转化前,MAD=3.36岁,Z-score转化后,MAD=3.42岁。焦磷酸测序法,65样本集数据Z-score转化前,MAD=4.20岁,Z-score转化后,MAD=2.76岁;62样本集数据Z-score转化前,MAD=3.92岁,Z-score转化后,MAD=3.63岁。
结论
Z-score转化方法能够有效的消除MassARRAY和焦磷酸测序数据之间的系统性批次效应;MassARRAY数据可以直接使用进行样本的年龄预测;使用焦磷酸测序数据进行年龄预测结果误差较大,但可以通过多样本积累进行Z-score转化之后预测年龄。
Keywords: 年龄推断, DNA甲基化, MassARRAY, 焦磷酸测序, 逐步多元线性回归模型, Z-score
Abstract
Objective
To study the difference in age estimation based on quantitative analysis of DNA methylation by MassARRAY and pyrosequencing techniques.
Methods
The methylation levels of 9 CpG sites from two independent whole blood sample sets (containing 65 and 62 samples) were detected using MassARRAY and pyrosequencing techniques. Z-score transformation was used to remove the batch effects of different techniques, and a linear regression model was used for age prediction.
Results
For age prediction using the MassARRAY system, the 65 samples showed a mean absolute difference (MAD) of 2.49 years before Z-score transformation of the data and 2.44 years after the transformation, similar to the results in the 62 samples (MAD of 3.36 years before and 3.42 years after Z-score transformation). For data typed from pyrosequencing, the 65 samples showed a MAD of 4.20 years before and 2.76 years after data Z-score transformation, also similar to the results in the 62 samples (MAD of 3.92 years before and 3.63 years after data transformation).
Conclusion
Z-score transformation can effectively reduce the system batch effect between MassARRAY and pyrosequencing. Data from the MassARRAY system allows direct age estimation without further data processing, while the pyrosequencing data may increase the error in age estimation, which can be corrected by Z-score transformation of the data.
Keywords: age estimation, DNA methylation, MassARRAY detection, Pyrosequencing, stepwise multivariate linear regression model, Z-score
年龄推断是法庭科学研究的重要内容,不仅可使用年龄作为线索直接缩小嫌疑人范围,而且可用于预测嫌疑人外貌<sup>[<xref ref-type="bibr" rid="b1">1</xref>-<xref ref-type="bibr" rid="b6">6</xref>]</sup>。DNA甲基化作为表观遗传学的重要组成部分,因其出色的准确性和稳定性成为年龄推断的研究热点<sup>[<xref ref-type="bibr" rid="b7">7</xref>-<xref ref-type="bibr" rid="b11">11</xref>]</sup>。越来越多的证据表明,人类基因组特定位点的DNA甲基化水平与年龄具有显著相关性,被称为“表观遗传学时钟”。近年来的研究发现,全基因组DNA甲基化在机体生长、发育、衰老的过程中存在着动态变化过程,基因组DNA甲基化总体水平随年龄增加而降低<sup>[<xref ref-type="bibr" rid="b12">12</xref>]</sup>,部分位点的甲基化水平随年龄增加而升高<sup>[<xref ref-type="bibr" rid="b13">13</xref>-<xref ref-type="bibr" rid="b14">14</xref>]</sup>。通过检测DNA甲基化变化,可以构建与之相关的年龄变化模型<sup>[<xref ref-type="bibr" rid="b15">15</xref>-<xref ref-type="bibr" rid="b16">16</xref>]</sup>,用于推断个体年龄。目前,检测DNA甲基化的方法<sup>[<xref ref-type="bibr" rid="b17">17</xref>-<xref ref-type="bibr" rid="b19">19</xref>]</sup>有很多,如SNaPshot <sup>[<xref ref-type="bibr" rid="b20">20</xref>]</sup>、Illumina阵列<sup>[<xref ref-type="bibr" rid="b21">21</xref>]</sup>、焦磷酸测序<sup>[<xref ref-type="bibr" rid="b22">22</xref>]</sup>、MassARRAY <sup>[<xref ref-type="bibr" rid="b16">16</xref>, <xref ref-type="bibr" rid="b23">23</xref>]</sup>等。由于每种方法之间的差异,测定的甲基化水平有所不同,因此基于某种方法开发的DNA甲基化年龄推断模型并不适用于另一种方法。本研究使用的年龄预测模型是我们实验室前期建立的模型,基于MassARRAY检测的390份北方汉族男性血液样本,使用9个甲基化位点构建的多元线性回归模型<sup>[<xref ref-type="bibr" rid="b23">23</xref>-<xref ref-type="bibr" rid="b24">24</xref>]</sup>。对于MassARRAY检测数据,可通过开发的<a href="http://liufan.big.ac.cn/AgePrediction/" target="_blank">http://liufan.big.ac.cn/AgePrediction/</a>网站直接进行年龄推断<sup>[<xref ref-type="bibr" rid="b23">23</xref>-<xref ref-type="bibr" rid="b24">24</xref>]</sup>。在我们的研究中发现,MassARRAY与焦磷酸测序获取的甲基化数据存在的系统性批次效应可以被Z-score转化部分消除,并能够有效降低利用焦磷酸测序检测数据进行年龄预测的误差。为了进一步验证我们的发现,最终建立一个跨方法的年龄推断模型,我们对65份和62份两个独立样本集进行了焦磷酸测序,进一步探究了两种方法在DNA甲基化年龄推断方面的差异,为DNA甲基化推断年龄应用于法庭科学以及案件侦查提供实验和理论基础。
1. 资料和方法
1.1. 样本收集与DNA提取
本实验室共收集127份无亲缘关系的中国北方汉族健康男性外周血样本,均无吸烟史,年龄分布在15~75岁之间。使用DNA提取试剂盒(QIAamp DNA Blood min kit,批号51106,Qiagen)提取基因组DNA。所有样本采集均已获得知情同意,所有样本均保存在-80 ℃冰箱。研究得到中华人民共和国公安部物证鉴定中心审查委员会的批准。
1.2. MassARRAY
使用EZ DNA甲基化转化试剂盒(Zymo Research)进行基因组DNA甲基化转化,起始DNA量为1 μg,产生20 µL转化后DNA,整个过程需要5.5 h。所有样本都经过PCR扩增、虾碱性磷酸酶(SAP)纯化、碱基特异性的酶切(T切),扩增体系:5 µL终体积中包括1 µL转化后DNA;1.37 µL ddH2O;0.5 µL 10×PCR Buffer;0.04 µL dNTP mix(25 mmol each);0.09 µL 5 U/µL PCR酶;2 µL 1 µmol Primer Mix(T7Reverse & 10 mer Forward),SAP纯化体系:1.7 µL ddH2O;0.3 µL 1.7 U/µLSAP,T切体系:3.15 µL RNase- free ddH2O;0.89 µL 5× T7 polymerase buffer;0.24 µL T cleavage mix;0.22 µL DTT(100 mmol);0.44 µL T7 RNA & DNA polymerase;0.06 µL RNase A。每个PCR反应通过以下程序在96微孔板上进行:95 ℃ 4 min→(95 ℃ 20 s、56 ℃ 30 s、72 ℃ 60 s)45个循环→72 ℃ 3 min→ 4 ℃ forever;37 ℃ 20 min→85 ℃ 5 min;37 ℃ 3 h→ 4 ℃ forever。树脂纯化之后,使用RS1000检测甲基化水平并通过MassARRAY EpiTYPER软件分析。
1.3. 焦磷酸测序
使用EpiTect Fast Bisulfite Kit(Qiagen)进行基因组DNA亚硫酸氢盐处理,取1 μg全基因组DNA,转化条件为95 ℃、5 min,60 ℃、10 min,95 ℃、5 min,60 ℃、10 min。PCR扩增体系为25 µL,包括0.5 µL重亚硫酸盐转化后DNA,12.5 µL 2 × PyroMark PCR Master Mix,2.5 µL 10×CoralLoad Concentrate,0.5 µL上游引物,0.5 µL下游引物,8.5 µL RNase-free water。热循环参数:扩增条件为95 ℃预变性15 min;94 ℃变性30 s,56 ℃退火30 s,72 ℃延伸30 s,共45个循环;72 ℃延伸10 min。将带有生物素标记的PCR产物与亲和素标记的微珠相混合,体系80 µL,包括5 µL PCR产物,40 µL PyroMark Binding,34 µL RNase-free water,1 µL Streptavidin Sepharose High Performance。利用PyroMark Q24 Advanced真空工作站分离纯化生物素标记的单链DNA。将单链的DNA模板与测序引物结合体系20 µL,包括0.75 μmol/L测序引物,19.25 µL PyroMark Annealing Buffer,80 ℃退火5 min,最终在PyroMark Q24 Advanced焦磷酸测序仪上进行测序。用PyroMark Q24 Advanced软件对数据进行分析。
1.4. MassARRAY和焦磷酸测序方法质控
MassARRAY方法的质控主要在3个方面:(1)使用Zymo Research转化试剂盒进行亚硫酸氢盐转化,Nanodrop2000进行定量,通过定量结果进行监测,建议1 μg新鲜的外周血DNA转化后定量浓度为20 ng/μL左右,A260/A280 > 2,本实验亚硫酸氢盐转化后ssDNA的浓度均高于20 ng/μL,A260/A280在2~2.2之间,我们认为符合要求;(2)MassARRAY设计引物时,引物中non-CpG‘C’最少数量设置默认值4,以评估亚硫酸氢盐处理是否完全,确保结果的可靠性;(3)在实验中使用了标准品或已知DNA甲基化值的DNA样品作为对照。
焦磷酸测序方法具有内置的质量控制,会在检测结果中给出亚硫酸氢盐转化程度质控,即任何一个其后没有G相连的C都被作为亚硫酸氢盐反应的质控,评估亚硫酸氢盐处理是否完全,并能预防假阳性甲基化检测,从而可确保结果的可靠性。实验过程中我们监测C峰不存在即为合格。
1.5. 位点及模型选择
本研究基于本实验室发表的利用MassARRAY检测数据筛选出的9个与年龄高度相关的DNA甲基化位点的多元线性回归年龄预测模型<sup>[<xref ref-type="bibr" rid="b23">23</xref>]</sup>。对于原始数据,可以通过我们开发的<a href="http://liufan.big.ac.cn/AgePrediction/" target="_blank">http://liufan.big.ac.cn/AgePrediction/</a>网站直接进行年龄推断。对于经过Z-score转化的数据,我们首先利用对390个基于MassARRAY的原始样本数据执行Z-score转化,并重新拟合年龄预测模型,随后再进行独立样本的年龄预测验证。
1.6. Z-score方法
![]() |
其中xij代表第i个甲基化位点在第j个样本中的信号值,xi和σi代表第i个甲基化位点在所有样本中信号的平均值和标准差,Zij代表经过转化后的信号值。
1.7. 统计学分析
缺失值通过K最近邻分类算法(KNN)进行补缺[25]。对于焦磷酸测序获得的9个甲基化位点的数据,首先利用对390个基于MassARRAY方法检测的原始样本数据执行Z-score转化,并重新拟合年龄预测模型,随后再进行独立样本的年龄预测验证。数据的Z-score转化、箱式图、散点图、Pearson相关系数、平均绝对误差(MAD)和拟合度(R2)的分析与计算都通过R语言实现(R版本号3.3.2)。P < 0.05为差异有统计学意义。
2. 结果
2.1. 9个CpG位点的甲基化水平数据在MassARRAY和焦磷酸测序之间存在系统性差异
利用MassARRAY和焦磷酸测序检测了65样本集和62样本集的9个甲基化位点的甲基化水平(图 1)。Pearson相关分析数据表明,两种方法对应的位点相关性较强(表 1、2)。结果显示,对于原始数据,两种方法检测的数据波动相对范围较大,显示了不同数据监测方法之间的系统性批次效应(图 1A、C)。数据经Z-score转化后,两种方法检测的甲基化水平被限制在相近的波动范围内,显示了系统性批次效应的消除(图 1B、D)。
1.

MassARRAY和焦磷酸测序的CpG位点的甲基化水平比较
Methylation levels of the CpG sites typed from MassARRAY and pyrosequencing data. A, B: Results in the 65 samples before and after Z-score transformation; C, D: Results in the 62 samples before and after Z-score transformation. The methylation levels typed from MassARRAY and pyrosequencing are represented in green and red, respectively.
1.
基于65样本集的MassARRAY和焦磷酸测序两种方法的相关性分析
Correlation analysis of MassARRAY and pyrosequencing based on 65 samples
| Factor | cg7 | cg12 | cg22 | cg83 | cg89 | cg104 | cg105 | cg111 | cg126 |
| *Pearson correlation coefficient. | |||||||||
| cg7 | 0.94* | 0.67 | -0.73 | -0.66 | 0.75 | 0.86 | 0.85 | -0.81 | 0.89 |
| cg12 | 0.73 | 0.81* | -0.46 | -0.27 | 0.63 | 0.73 | 0.72 | -0.55 | 0.57 |
| cg22 | -0.74 | -0.46 | 0.94* | 0.69 | -0.58 | -0.72 | -0.75 | 0.78 | -0.86 |
| cg83 | -0.74 | -0.27 | 0.76 | 0.91* | -0.45 | -0.69 | -0.63 | 0.63 | -0.82 |
| cg89 | 0.79 | 0.75 | -0.53 | -0.46 | 0.94* | 0.73 | 0.77 | -0.65 | 0.75 |
| cg104 | 0.8 | 0.67 | -0.66 | -0.67 | 0.69 | 0.9* | 0.85 | -0.68 | 0.81 |
| cg105 | 0.86 | 0.65 | -0.68 | -0.73 | 0.75 | 0.86 | 0.9* | -0.78 | 0.86 |
| cg111 | -0.86 | -0.6 | 0.74 | 0.6 | -0.71 | -0.76 | -0.81 | 0.96* | -0.87 |
| cg126 | 0.86 | 0.56 | -0.81 | -0.76 | 0.68 | 0.81 | 0.8 | -0.83 | 0.95* |
2.
基于62样本集的MassARRAY和焦磷酸测序两种方法的相关性分析
Correlation analysis of MassARRAY and pyrosequencing based on 62 samples
| Factor | cg7 | cg12 | cg22 | cg83 | cg89 | cg104 | cg105 | cg111 | cg126 |
| *Pearson correlation coefficient. | |||||||||
| cg7 | 0.49* | 0.41 | -0.41 | -0.5 | 0.3 | 0.22 | 0.41 | -0.15 | 0.44 |
| cg12 | 0.66 | 0.82* | -0.35 | -0.36 | 0.43 | 0.24 | 0.29 | -0.33 | 0.51 |
| cg22 | 0.1 | 0.12 | 0.08* | -0.08 | 0.12 | 0.31 | 0.2 | -0.1 | 0.18 |
| cg83 | -0.73 | -0.4 | 0.73 | 0.81* | -0.52 | -0.42 | -0.55 | 0.49 | -0.75 |
| cg89 | 0.63 | 0.56 | -0.5 | -0.54 | 0.76* | 0.41 | 0.55 | -0.35 | 0.69 |
| cg104 | 0.78 | 0.57 | -0.71 | -0.66 | 0.52 | 0.47* | 0.52 | -0.56 | 0.81 |
| cg105 | 0.32 | 0.31 | -0.28 | -0.28 | 0.15 | 0.16 | 0.21* | -0.04 | 0.22 |
| cg111 | -0.73 | -0.5 | 0.73 | 0.64 | -0.49 | -0.28 | -0.56 | 0.63* | -0.77 |
| cg126 | 0.69 | 0.46 | -0.59 | -0.58 | 0.45 | 0.3 | 0.49 | -0.52 | 0.75* |
2.2. 65样本集和62样本集经Z-score转化前后年龄预测的线性拟合图
在分析过程中,我们在65样本集上观测到了与62样本集相似的结果。对于MassARRAY监测的数据,不论是否经过Z-score转化,其年龄预测的MAD基本没有变化。对65样本集的MassARRAY数据,Z-score转换之前预测结果的MAD=2.49(图 2A),Z-score转换之后的预测结果为MAD=2.44(图 2C);对于62样本集的MassARRAY数据,Z-score转换之前预测结果的MAD=3.36(图 3A),Z-score转换之后的预测结果为MAD=3.42(图 3C)。
2.

预测年龄(y轴)与真实年龄(x轴)散点图
Scatter plot of predicted age against chronological age in the 65 samples. A, B: Plots based on data without Z-score transformation typed from MassARRAY and pyrosequencing. C, D: Plots based on data after Z-score transformation typed from MassARRAY and pyrosequencing.
3.

预测年龄(y轴)与真实年龄(x轴)散点图
Scatter plot of predicted age against chronological age in the 62 samples. A, B: Plots based on data without Z-score transformation typed from MassARRAY and pyrosequencing. C, D: Plots based on data after Z-score transformation typed from MassARRAY and pyrosequencing.
但是,对于焦磷酸测序获得的数据,其年龄预测的MAD在Z-score转化前后变化很大。对65样本集,Z-score转换之前预测结果的MAD=4.20(图 2B),Z-score转换之后的预测结果为MAD=2.76(图 2D);对于62样本集的MassARRAY数据,Z-score转换之前预测结果的MAD=3.92(图 3B),Z-score转换之后的预测结果为MAD=3.63(图 3D)。
3. 讨论
本研究是在本实验室已经发表的9位点DNA甲基化年龄推断模型研究的基础上,开展旨在探讨高通量的MassARRAY技术和焦磷酸测序技术在DNA甲基化年龄推断中对推断结果的准确性的影响及差异比较研究,为后续建立跨方法的年龄推断模型奠定基础。
甲基化的定量检测分为3步:亚硫酸氢盐转化、PCR扩增、甲基化测定。MassARRAY和焦磷酸测序在技术原理方面存在差异,其差异重点是第三步甲基化测定。首先检测前对基因组DNA进行亚硫酸氢盐转化[26],DNA中未甲基化的胞嘧啶(C)转变为尿嘧啶(U),而甲基化的胞嘧啶(C)保持不变,因此将甲基化的差异变为序列的差异。亚硫酸氢盐转换后进行PCR扩增,将序列差异性成比例的放大后用于下一步的甲基化定量检测。在甲基化测定中,MassARRAY和焦磷酸测序技术均需要4~5个小时完成定量检测,而且都被用于法庭科学年龄推断研究[16, 27],但是两种方法检测原理不同。MassARRAY结合了碱基特异性酶切反应和MALDI-TOF检测原理,可实现多个CpG的分析检测,碱基特异性酶切实验由亚硫酸氢盐处理待测DNA开始,经过亚硫酸氢盐处理,由此在DNA模板中产生甲基化特异的序列变化;利用5'末端带有T7-启动子的引物进行PCR扩增,产物经SAP处理后用于碱基特异性的酶切反应;酶切后DNA片段的大小和分子量取决于亚硫酸氢盐处理后的碱基变化,飞行质谱能测出每个片段的分子量,配套软件MassARRAY EpiTYPER则能自动报告每个相应片段的甲基化程度[27-28]。焦磷酸测序是一种新的DNA序列分析技术,是在同一反应体系中由4中酶催化的酶级联化学发光反应。特异性引物与模板DNA退火后,在DNA聚合酶、ATP硫酸化酶、荧光素酶和三磷酸腺苷双磷酸酶4种酶的协同作用下,将每一个dNTP的聚合与一次荧光信号的释放偶联起来,荧光信号的强度与结合的核苷酸数目成比例,然后通过荧光检测装置检测荧光的释放和强度,并通过软件将光信号转换为峰图。反应底物为5-磷酰硫酸和荧光素,反应体系包括待测序DNA单链和测序引物。每一轮测序反应中,加入1种dNTP,若该dNTP与模板配对,聚合酶可将其掺入到引物链中并释放出等摩尔数的焦磷酸集团;硫酸化酶催化APS和PPi形成ATP,后者驱动荧光素酶介导的荧光素向氧化荧光素转化,发出与ATP量成正比的可见光信号,并由Pyrogram转化为一个峰值,其高度与反应中掺入的核苷酸数目成正比;根据加入dNTP类型和荧光信号强度就可实时记录模板DNA的核苷酸序列[29-30]。在本研究中,选取了两个批次且相互独立的样本,分别使用MassARRAY和焦磷酸测序对9个甲基化位点进行检测,并进行了个体年龄推断分析。结果显示,进行Z-score转换前后,MassARRAY测得的MAD值差异不明显;焦磷酸测序数据由MAD=4.20年降低为MAD=2.76年,可以更加准确的预测年龄,因此,Z-score转换能够有效降低MassARRAY和焦磷酸测序数据之间的系统性批次效应。
本项研究进一步证明不同的甲基化检测方法存在系统性差异,而且这种差异可以通过Z-score转换算法得到有效降低。通过Z-score转换技术可以有效的将利用焦磷酸测序DNA甲基化数据进行年龄预测的误差降低到同MassARRAY相似的水平。但是,由于目前的年龄预测模型是基于MassARRAY数据构建的,因此对于MassARRAY检测的数据,可以直接通过我们已发表的年龄预测网站进行年龄推断(<a href="http://liufan.big.ac.cn/AgePrediction/" target="_blank">http://liufan.big.ac.cn/AgePrediction/</a>)<sup>[<xref ref-type="bibr" rid="b23">23</xref>-<xref ref-type="bibr" rid="b24">24</xref>]</sup>,进而应用于刑事科学技术年龄推断;焦磷酸测序数据则必须首先将检测数据进行Z-score转化才能得到相对准确的预测结果,而单个样本无法进行Z-score转化以及样本太少会造成较大的Z-score转化误差,因此在利用Z-score转化模型时需要时间积累焦磷酸平台样本数据,但这对于实际的案件应用所面临的迫切性是矛盾的。针对这个问题,我们建议目前统一使用MassARRAY进行DNA甲基化水平检测。另外,随着时间的积累,当积累了足够多的焦磷酸测序数据后,重新训练一个针对焦磷酸测序检测方法的年龄预测模型不失为一个更好的解决方案,并对比两种方法的年龄预测成效,从而使甲基化推断年龄在刑事技术方面发挥更大的作用。
Biographies
王玲,硕士,E-mail: ailinandfriend@163.com
彭付端,博士,E-mail: pengfuduan@big.ac.cn
Funding Statement
国家重点研发计划资助项目(2017YFC0803501);基本科研业务费课题(2016JB037);公安部技术研究计划项目(2016JSYJA04)
Contributor Information
王 玲 (Ling WANG), Email: ailinandfriend@163.com.
彭 付端 (Fuduan PENG), Email: pengfuduan@big.ac.cn.
刘 天资 (Tianzi LIU), Email: liutz@big.ac.cn.
丰 蕾 (Lei FENG), Email: fengleink@163.com.
References
- 1.Park JL, Kim JH, Seo E, et al. Identification and evaluation of agecorrelated DNA methylation markers for forensic use. http://europepmc.org/abstract/MED/27017110. Forensic Sci Int Genet. 2016;23:64–70. doi: 10.1016/j.fsigen.2016.03.005. [Park JL, Kim JH, Seo E, et al. Identification and evaluation of agecorrelated DNA methylation markers for forensic use[J]. Forensic Sci Int Genet, 2016, 23: 64-70.] [DOI] [PubMed] [Google Scholar]
- 2.Márquez-Ruiz AB, González-Herrera L, Luna JD, et al. DNA methylation levels and telomere length in human teeth: usefulness for age estimation. http://link.springer.com/article/10.1007/s00414-019-02242-7. Int J Legal Med. 2020;134(2):451–9. doi: 10.1007/s00414-019-02242-7. [Márquez-Ruiz AB, González-Herrera L, Luna JD, et al. DNA methylation levels and telomere length in human teeth: usefulness for age estimation[J]. Int J Legal Med, 2020, 134(2): 451-9.] [DOI] [PubMed] [Google Scholar]
- 3.Pan C, Yi SH, Xiao C, et al. The evaluation of seven age-related CPGS for forensic purpose in blood from Chinese Han population. http://www.researchgate.net/publication/338807303_The_evaluation_of_seven_age-related_CpGs_for_forensic_purpose_in_blood_from_Chinese_Han_population. Forensic Sci Int Genet. 2020;46:102251. doi: 10.1016/j.fsigen.2020.102251. [Pan C, Yi SH, Xiao C, et al. The evaluation of seven age-related CPGS for forensic purpose in blood from Chinese Han population [J]. Forensic Sci Int Genet, 2020, 46: 102251.] [DOI] [PubMed] [Google Scholar]
- 4.Kayser M. Forensic DNA Phenotyping: Predicting human appearance from crime scene material for investigative purposes. http://www.tandfonline.com/servlet/linkout?suffix=CIT0025&dbid=8&doi=10.1080%2F14636778.2018.1469975&key=25716572. Forensic Sci Int Genet. 2015;18:33–48. doi: 10.1016/j.fsigen.2015.02.003. [Kayser M. Forensic DNA Phenotyping: Predicting human appearance from crime scene material for investigative purposes[J]. Forensic Sci Int Genet, 2015, 18: 33-48.] [DOI] [PubMed] [Google Scholar]
- 5.Yi SH, Jia YS, Mei K, et al. Age-related DNA methylation changes for forensic age-prediction. http://europepmc.org/abstract/med/25399049. Int J Legal Med. 2015;129(2):237–44. doi: 10.1007/s00414-014-1100-3. [Yi SH, Jia YS, Mei K, et al. Age-related DNA methylation changes for forensic age-prediction[J]. Int J Legal Med, 2015, 129(2): 237- 44.] [DOI] [PubMed] [Google Scholar]
- 6.Huang Y, Yan J, Hou JY, et al. Developing a DNA methylation assay for human age prediction in blood and bloodstain. http://www.sciencedirect.com/science/article/pii/S1872497315300119. Forensic Sci Int Genet. 2015;17:129–36. doi: 10.1016/j.fsigen.2015.05.007. [Huang Y, Yan J, Hou JY, et al. Developing a DNA methylation assay for human age prediction in blood and bloodstain[J]. Forensic Sci Int Genet, 2015, 17: 129-36.] [DOI] [PubMed] [Google Scholar]
- 7.Lee HY, Jung SE, Oh YN, et al. Epigenetic age signatures in the forensically relevant body fluid of semen: a preliminary study. http://www.sciencedirect.com/science/article/pii/S1872497315300181. Forensic Sci Int Genet. 2015;19:28–34. doi: 10.1016/j.fsigen.2015.05.014. [Lee HY, Jung SE, Oh YN, et al. Epigenetic age signatures in the forensically relevant body fluid of semen: a preliminary study[J]. Forensic Sci Int Genet, 2015, 19: 28-34.] [DOI] [PubMed] [Google Scholar]
- 8.Correia Dias H, Cordeiro C, Corte Real F, et al. Age estimation based on DNA methylation using blood samples from deceased individuals. http://onlinelibrary.wiley.com/doi/10.1111/1556-4029.14185/abstract. J Forensic Sci. 2020;65(2):465–70. doi: 10.1111/1556-4029.14185. [Correia Dias H, Cordeiro C, Corte Real F, et al. Age estimation based on DNA methylation using blood samples from deceased individuals[J]. J Forensic Sci, 2020, 65(2): 465-70.] [DOI] [PubMed] [Google Scholar]
- 9.王 中华, 李 淑瑾. DNA甲基化推断人类年龄的研究进展. 法医学杂志. 2020;36(1):109–14, 119. [王中华, 李淑瑾. DNA甲基化推断人类年龄的研究进展[J].法医学杂志, 2020, 36(1): 109-14, 119.] [Google Scholar]
- 10.Weidner CI, Lin Q, Koch CM, et al. Aging of blood can be tracked by DNA methylation changes at just three CpG sites. http://pubmedcentralcanada.ca/pmcc/articles/PMC4053864/?report=classic. Genome Biol. 2014;15(2):R24. doi: 10.1186/gb-2014-15-2-r24. [Weidner CI, Lin Q, Koch CM, et al. Aging of blood can be tracked by DNA methylation changes at just three CpG sites[J]. Genome Biol, 2014, 15(2): R24.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.赵 书民, 李 成涛. DNA甲基化在法医学中的应用前景及其检测方法新进展. http://d.wanfangdata.com.cn/periodical/fyxzz200904014. 法医学杂志. 2009;25(4):290–5. [赵书民, 李成涛. DNA甲基化在法医学中的应用前景及其检测方法新进展[J].法医学杂志, 2009, 25(4): 290-5.] [PubMed] [Google Scholar]
- 12.Wilson VL, Jones PA. DNA methylation decreases in aging but not in immortal cells. http://www.researchgate.net/publication/16353945_Wilson_VL_Jones_PA_DNA_methylation_decreases_in_aging_but_not_in_immortal_cells_Science_220_1055-1057. Science. 1983;220(4601):1055–7. doi: 10.1126/science.6844925. [Wilson VL, Jones PA. DNA methylation decreases in aging but not in immortal cells[J]. Science, 1983, 220(4601): 1055-7.] [DOI] [PubMed] [Google Scholar]
- 13.Rakyan VK, Down TA, Maslau S, et al. Human aging-associated DNA hypermethylation occurs preferentially at bivalent chromatin domains. http://dx.doi.org/10.1101/gr.103101.109. Genome Res. 2010;20(4):434–9. doi: 10.1101/gr.103101.109. [Rakyan VK, Down TA, Maslau S, et al. Human aging-associated DNA hypermethylation occurs preferentially at bivalent chromatin domains[J]. Genome Res, 2010, 20(4): 434-9.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Beerman I, Bock C, Garrison BS, et al. Proliferation-dependent alterations of the DNA methylation landscape underlie hematopoietic stem cell aging. http://www.sciencedirect.com/science/article/pii/S1934590913000209. Cell Stem Cell. 2013;12(4):413–25. doi: 10.1016/j.stem.2013.01.017. [Beerman I, Bock C, Garrison BS, et al. Proliferation-dependent alterations of the DNA methylation landscape underlie hematopoietic stem cell aging[J]. Cell Stem Cell, 2013, 12(4): 413-25.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zbieć-Piekarska R, Spólnicka M, Kupiec T, et al. Development of a forensically useful age prediction method based on DNA methylation analysis. http://www.sciencedirect.com/science/article/pii/S1872497315300041. Forensic Sci Int Genet. 2015;17:173–9. doi: 10.1016/j.fsigen.2015.05.001. [Zbieć-Piekarska R, Spólnicka M, Kupiec T, et al. Development of a forensically useful age prediction method based on DNA methylation analysis[J]. Forensic Sci Int Genet, 2015, 17: 173-9.] [DOI] [PubMed] [Google Scholar]
- 16.Freire-Aradas A, Phillips C, Mosquera-Miguel A, et al. Development of a methylation marker set for forensic age estimation using analysis of public methylation data and the Agena Bioscience EpiTYPER system. http://www.ncbi.nlm.nih.gov/pubmed/27337627. Forensic Sci Int Genet. 2016;24:65–74. doi: 10.1016/j.fsigen.2016.06.005. [Freire-Aradas A, Phillips C, Mosquera-Miguel A, et al. Development of a methylation marker set for forensic age estimation using analysis of public methylation data and the Agena Bioscience EpiTYPER system[J]. Forensic Sci Int Genet, 2016, 24: 65-74.] [DOI] [PubMed] [Google Scholar]
- 17.周 娣, 王 正, 张 霁, et al. DNA甲基化标记检测方法及其法医学应用. http://www.cnki.com.cn/Article/CJFDTotal-FUAN201305011.htm. 中国法医学杂志. 2013;28(5):394–7. [周娣, 王正, 张霁, 等. DNA甲基化标记检测方法及其法医学应用[J].中国法医学杂志, 2013, 28(5): 394-7.] [Google Scholar]
- 18.聂 燕钗, 俞 丽娟, 管 桦, et al. DNA甲基化检测方法及其法医学应用研究进展. http://d.wanfangdata.com.cn/periodical/fyxzz201703017. 法医学杂志. 2017;33(3):293–300. doi: 10.3969/j.issn.1004-5619.2017.03.017. [聂燕钗, 俞丽娟, 管桦, 等. DNA甲基化检测方法及其法医学应用研究进展[J].法医学杂志, 2017, 33(3): 293-300.] [DOI] [PubMed] [Google Scholar]
- 19.Xu HM, Zhao Y, Liu ZP, et al. Bisulfite genomic sequencing of DNA from dried blood spot microvolume samples. http://www.ncbi.nlm.nih.gov/pubmed/21737370. Forensic Sci Int Genet. 2012;6(3):306–9. doi: 10.1016/j.fsigen.2011.06.007. [Xu HM, Zhao Y, Liu ZP, et al. Bisulfite genomic sequencing of DNA from dried blood spot microvolume samples[J]. Forensic Sci Int Genet, 2012, 6(3): 306-9.] [DOI] [PubMed] [Google Scholar]
- 20.Jung SE, Lim SM, Hong SR, et al. DNA methylation of the ELOVL2, FHL2, KLF14, C1orf132/MIR29B2C, and TRIM59 genes for age prediction from blood, saliva, and buccal swab samples. http://www.ncbi.nlm.nih.gov/pubmed/30300865. Forensic Sci Int Genet. 2019;38:1–8. doi: 10.1016/j.fsigen.2018.09.010. [Jung SE, Lim SM, Hong SR, et al. DNA methylation of the ELOVL2, FHL2, KLF14, C1orf132/MIR29B2C, and TRIM59 genes for age prediction from blood, saliva, and buccal swab samples[J]. Forensic Sci Int Genet, 2019, 38: 1-8.] [DOI] [PubMed] [Google Scholar]
- 21.Alsaleh H, Haddrill PR. Identifying blood-specific age-related DNA methylation markers on the Illumina MethylationEPICⓇ BeadChip. http://www.sciencedirect.com/science/article/pii/S0379073819303561. Forensic Sci Int. 2019;303:109944. doi: 10.1016/j.forsciint.2019.109944. [Alsaleh H, Haddrill PR. Identifying blood-specific age-related DNA methylation markers on the Illumina MethylationEPICⓇ BeadChip [J]. Forensic Sci Int, 2019, 303: 109944.] [DOI] [PubMed] [Google Scholar]
- 22.Li LY, Song F, Lang M, et al. Methylation-based age prediction using pyrosequencing platform from seminal stains in Han Chinese males. http://www.researchgate.net/publication/335707268_Methylation-Based_Age_Prediction_Using_Pyrosequencing_Platform_from_Seminal_Stains_in_Han_Chinese_Males. J Forensic Sci. 2020;65(2):610–9. doi: 10.1111/1556-4029.14186. [Li LY, Song F, Lang M, et al. Methylation-based age prediction using pyrosequencing platform from seminal stains in Han Chinese males[J]. J Forensic Sci, 2020, 65(2): 610-9.] [DOI] [PubMed] [Google Scholar]
- 23.Feng L, Peng FD, Li SF, et al. Systematic feature selection improves accuracy of methylation-based forensic age estimation in Han Chinese males. http://www.ncbi.nlm.nih.gov/pubmed/29631189. Forensic Sci Int Genet. 2018;35:38–45. doi: 10.1016/j.fsigen.2018.03.009. [Feng L, Peng FD, Li SF, et al. Systematic feature selection improves accuracy of methylation-based forensic age estimation in Han Chinese males[J]. Forensic Sci Int Genet, 2018, 35: 38-45.] [DOI] [PubMed] [Google Scholar]
- 24.Peng FD, Feng L, Chen J, et al. Validation of methylation-based forensic age estimation in time-series bloodstains on FTA cards and gauze at room temperature conditions. http://www.ncbi.nlm.nih.gov/pubmed/30878720. Forensic Sci Int Genet. 2019;40:168–74. doi: 10.1016/j.fsigen.2019.03.006. [Peng FD, Feng L, Chen J, et al. Validation of methylation-based forensic age estimation in time-series bloodstains on FTA cards and gauze at room temperature conditions[J]. Forensic Sci Int Genet, 2019, 40: 168-74.] [DOI] [PubMed] [Google Scholar]
- 25.Zhang SC, Li XL, Zong M, et al. Efficient kNN classification with different numbers of nearest neighbors. http://ieeexplore.ieee.org/document/7898482. IEEE Trans Neural Netw Learn Syst. 2018;29(5):1774–85. doi: 10.1109/TNNLS.2017.2673241. [Zhang SC, Li XL, Zong M, et al. Efficient kNN classification with different numbers of nearest neighbors[J]. IEEE Trans Neural Netw Learn Syst, 2018, 29(5): 1774-85.] [DOI] [PubMed] [Google Scholar]
- 26.路 兆宁, 周 在威, 马 晴雯, et al. 3种DNA甲基化定量分析方法的比较. http://d.wanfangdata.com.cn/Periodical/gwyx-fzsw201502005. 医学分子生物学杂志. 2015;(2):84–9. [路兆宁, 周在威, 马晴雯, 等. 3种DNA甲基化定量分析方法的比较[J].医学分子生物学杂志, 2015(2): 84-9.] [Google Scholar]
- 27.Kunze S. Quantitative region-specific DNA methylation analysis by the EpiTYPER TM technology. http://www.springerlink.com/content/fulltext.pdf?id=doi:10.1007/978-1-4939-7481-8_26. Methods Mol Biol. 2018;1708:515–35. doi: 10.1007/978-1-4939-7481-8_26. [Kunze S. Quantitative region-specific DNA methylation analysis by the EpiTYPER TM technology[J]. Methods Mol Biol, 2018, 1708: 515-35.] [DOI] [PubMed] [Google Scholar]
- 28.Suchiman HE, Slieker RC, Kremer D, et al. Design, measurement and processing of region-specific DNA methylation assays: the mass spectrometry-based method EpiTYPER. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4585020/ Front Genet. 2015;6:287. doi: 10.3389/fgene.2015.00287. [Suchiman HE, Slieker RC, Kremer D, et al. Design, measurement and processing of region-specific DNA methylation assays: the mass spectrometry-based method EpiTYPER[J]. Front Genet, 2015, 6: 287.] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Busato F, Dejeux E, El Abdalaoui H, et al. Quantitative DNA methylation analysis at single-nucleotide resolution by pyrosequencingⓇ. http://link.springer.com/10.1007/978-1-4939-7481-8_22. Methods Mol Biol. 2018;1708:427–45. doi: 10.1007/978-1-4939-7481-8_22. [Busato F, Dejeux E, El Abdalaoui H, et al. Quantitative DNA methylation analysis at single-nucleotide resolution by pyrosequencingⓇ[J]. Methods Mol Biol, 2018, 1708: 427-45.] [DOI] [PubMed] [Google Scholar]
- 30.Busato F, Tost J. SNP-based quantification of allele-specific DNA methylation patterns by pyrosequencingⓇ. http://link.springer.com/10.1007/978-1-4939-2715-9_21. Methods Mol Biol. 2015;1315:291–313. doi: 10.1007/978-1-4939-2715-9_21. [Busato F, Tost J. SNP-based quantification of allele-specific DNA methylation patterns by pyrosequencingⓇ[J]. Methods Mol Biol, 2015, 1315: 291-313.] [DOI] [PubMed] [Google Scholar]

