建立并验证25个祖源信息标记物的集合用于族群信息推断  

Development and Performance Evaluation of a Novel Panel of 25 Ancestry Informative Markers for Forensic Ancestry Inference

作　　者：李耀庭 LI Yaoting(Department of Forensic Science,Guangdong Police College,Guangzhou,Guangdong 510440,China)

机构地区：[1]广东警官学院刑事技术系,广东广州510440

出　　处：《中国刑警学院学报》2023年第5期122-128,共7页Journal of Criminal Investigation Police University of China

基　　金：2021年度国家自然科学基金青年项目(编号:82000519);2021年度广东省普通高校青年创新人才项目(编号:2021KQNCX058);2022年度广州市基础研究与应用基础研究项目(编号:202201011266);2023年度广东省教育科学规划课题高等教育专项(编号:2023GXJK414)。

摘　　要：当传统的常染色体STR分型比对模式无法提供线索时,对犯罪现场的生物样本的祖源信息推断可为侦查指明方向。通过祖源信息标记物的集合对未知样本的族群信息推断,并验证模型预测能力。以朴素贝叶斯分类器算法筛选特征性的SNPs,建立一个包含25个祖源信息标记物集合Li-25AIMset,用于推断未知个体的大陆祖源信息。UMAP和STRUCTURE分析证实了Li-25AIMset模型的有效性,表明该模型能准确识别样本的祖源属性。Li-25AIMset与已发表的祖源信息标记物集合Kidd-55AIMset相比,其可通过检测更少的标记物,达到更高的准确性。通过人工智能算法对高通量测序数据挖掘,能筛选出高信息密度的标记物,有望应用在个体识别和表型预测,为侦查提供更多线索。When traditionalautosomalShort Tandem Repeat(STR)profiling is unable to provide valuable information to apprehend a criminal,forensic ancestry inference of biological samples left at the crime scene is likely to offer investigative leads and improve the efficiency of the investigation process.Consequently,we explored the ancestry informative markers set for inferring population information of unknown samples and validated the predictive capabilities of this model.A 25-plex ancestry informative markers set(Li-25AIMset)has been developed by assembling well-differentiated single nucleotide polymorphisms(SNPs)in nextgeneration sequencing data for ancestry assignment of unknown individuals from four continental populations,namely African,European,South Asian,and East Asian.The effectiveness of the Li-25AIMset was confirmed through UMAP analysis and STRUCTURE analysis,which demonstrated that it was capable of determining the ancestry origin of an unknown individual from the aforementioned populations.Additionally,a comparison between the Li-25AIMset and a published AIMs set(Kidd-55AIMset)revealed that the former,which utilized fewer detecting markers,displayed greater accuracy.The artificial-intelligence algorithm utilized in nextgeneration sequencing data has the potential to exploit more informative SNPs for individual identification and phenotype prediction,thus providing as many investigative clues as possible in the future.

关 键 词：祖源信息标志物 高通量测序 人工智能 单核苷酸多态性 

分 类 号：D919.1[医药卫生—法医学]