MPS's Key Laboratory of Forensic Genetics, National Engineering Laboratory for Crime Scene Evidence, Investigation and Examination, Institute of Forensic Science, Ministry of Public Security (MPS), Beijing 100038, China; Faculty of Forensic Sciences, Shanxi Medical University, Taiyuan 030001, Shanxi, China.
Chinese Center For Disease Control And Prevention, State Key Laboratory of Infectious Disease Prevention and Control, Beijing 102206, China.
Forensic Sci Int Genet. 2020 Sep;48:102337. doi: 10.1016/j.fsigen.2020.102337. Epub 2020 Jun 30.
Peripheral blood, menstrual blood, semen, saliva and vaginal secretions are the five most common body fluids found at crime scenes, and the identification of these five body fluids is of great significance to the reconstruction of a crime scene and resolution of the case. However, accurate identification of these five body fluids is still a challenge. To address this problem, a mathematical model for differentiating five types of forensic body fluids based on the differential expression characteristics of multiple miRNAs in five body fluids (peripheral blood, menstrual blood, semen, saliva and vaginal secretions) was developed. A total of 350 forensic body fluids (70 of each type) were collected and tested, and relative expression of 10 miRNAs (miR-451a, miR-205-5p, miR-203-3p, miR-214-3p, miR-144-3p, miR-144-5p, miR-654-5p, miR-888-5p, miR-891a-5p, miR-124a-3p) in all samples was detected by SYBR Green real-time qPCR. Three hundred samples (60 samples of each body fluid) were used as the training set to screen meaningful identification markers by stepwise discriminant analysis, and a discriminant function was established. Fifty samples (10 samples of each body fluid) were used as a validation set to examine the accuracy of the model, and 25 samples (the types of samples were unknown to the experimenter) were used for a blind test. Except for miR-144-3p, the other miRNAs were selected to construct discriminant analysis models. The self-validation accuracy of the model was 99.7 %, cross-validation accuracy was 99.3 %, accuracy of the identification validation set was 100 %, and accuracy of the blind test result was 100 %. This study provides a reliable and accurate identification strategy for five common body fluids (peripheral blood, menstrual blood, semen, saliva, and vaginal secretions) in forensic medicine.
外周血、月经血、精液、唾液和阴道分泌物是犯罪现场最常见的五种体液，鉴定这五种体液对重建犯罪现场和案件侦破具有重要意义。然而，准确鉴定这五种体液仍然是一个挑战。为了解决这个问题，我们建立了一种基于五种体液（外周血、月经血、精液、唾液和阴道分泌物）中多种 miRNA 差异表达特征的区分五种法医体液的数学模型。共收集和检测了 350 份法医体液（每种类型 70 份），通过 SYBR Green 实时 qPCR 检测所有样本中 10 种 miRNA（miR-451a、miR-205-5p、miR-203-3p、miR-214-3p、miR-144-3p、miR-144-5p、miR-654-5p、miR-888-5p、miR-891a-5p、miR-124a-3p）的相对表达量。以 300 份样本（每种体液 60 份）作为训练集，通过逐步判别分析筛选有意义的鉴别标志物，并建立判别函数。以 50 份样本（每种体液 10 份）作为验证集，检验模型的准确性，并对 25 份未知类型的样本进行盲测。除了 miR-144-3p 之外，其他 miRNA 都被用来构建判别分析模型。模型的自我验证准确率为 99.7%，交叉验证准确率为 99.3%，验证集的识别准确率为 100%，盲测结果的准确率为 100%。本研究为法医鉴定中五种常见体液（外周血、月经血、精液、唾液和阴道分泌物）提供了一种可靠、准确的鉴定策略。
