微单倍型测序在混合检材亲缘关系分析中的应用研究

目前，法医实践中常规的混合斑检验主要通过分析和检测检材的DNA遗传标记并与嫌疑人进行比对，或在法医DNA数据库中进行检索，以达到个人识别的目的。随着社会多元化发展和鉴定案情的复杂化，涉及亲缘关系鉴定分析的混合斑检验发挥着越来越重要的作用。现有基于PCR-CE技术的STR分型在解决这一难题时依然困难重重，经常不能获得明确的鉴定结果。本项目拟基于下一代测序技术的最新进展，筛选和构建一个高信息量微单倍型遗传标记的复合检测体系，为混合检材分析提供高效能的检测手段。同时，以三种常见亲缘关系为代表，应用贝叶斯网络等多种概率模型，对不同检验假设、不同类型的混合检材亲缘关系分析所获得的大信息量检测数据进行评估。基于实际检测数据和计算机模拟数据，结合等位基因丢失、插入、突变等因素，逐步建立和优化一个从混合检材微单倍型分型到亲缘关系证据评估的可执行策略，以满足公共安全领域日益复杂的鉴定需要。

Nowadays, the routine examination of DNA mixtures collected in forensic crime scene provide the essential clues for the investigation mainly by comparing the STR results of the mixture with the suspect, or searching in the forensic DNA database. With the development of social diversity and the increasing complexity in forensic practice, kinship analysis plays an increasingly important role in forensic practice. In this situation, the existing STR typing approach based on PCR-CE platform is still difficult to solve this problem and often fail to draw a clear conclusion. Based on the latest development of next generation sequencing technology, this study aims to screen and construct a multiplexing microhaplotype-based detecting system, which is expected to provide high detecting efficiency for DNA mixture analysis. At the same time, based on three kinds of common relationship, the Bayesian network and other probability models are used to analysis the microhaplotype sequencing data of mixture with different mixing proportion and condributors, and evaluate the weight of evidence under different testing assumptions, combining multiple essential factors, such as allele dropout, drop-in, mutation, et al. This study is expected to establish an optimized strategy for the relationship inference involving forensic DNA mixtures.