Identification of contributors in the low-template and mixed samples by forensic mathematical estimation and experimental analysis

分析混合或/和低模板DNA样本是当前法医DNA测试中具有挑战性但又不可避免的。我们重新评估常染色体STR多重分型试剂盒，发现杂合子等位基因峰值不平衡甚至发生在一个人的样本中。因此，杂合组合的测定并不总是可靠的，不使用等位基因峰高的二元法在2个或3个个体的混合样品中仍然显示出很强的效力。虽然人们常说，在混合样本中存在不能排除非贡献者的风险，但使用大量混合样本的模拟分析显示，贡献者的一级亲属被错误地包括为贡献者，尽管其比例很小。此外，对37个常染色体STR基因座进行了分型，估计了日本人群的等位基因频率和鉴别能力。最后，建立了混合样品的分析方法。

Analyzing mixed or/and low-template DNA samples is challenging but inevitable in current forensic DNA test. We reevaluated autosomal STR multiplex typing kit and found that allele peak imbalance of heterozygote occurred even in a single person sample. Therefore, the determination of the heterozygous combination is not always reliable and the binary method which does not use the allelic peak heights still showed very powerful in mixed samples from 2 or 3 individuals.While it is often said that there is a risk that non-contributors cannot be excluded in a mixed sample, the simulated analysis using a large number of the mixed samples revealed that the first grade kin of the contributor is erroneously included as a contributor although its proportion is small. Moreover, 37 loci of autosomal STR were genotyped and the allele frequencies and the discriminating power were estimated in the Japanese population. Finally, the analytical procedure from mixed samples has been established.