SMART:自主研发的混合STR图谱分析系统

SMART:an Analysis System for Mixed STR Profiles

目的混合STR图谱的分析是法医遗传学领域的一个难题。目前国内大多数法医DNA实验室依靠人工方法分析混合STR图谱,费时费力,且拆分效果也多有主观性,难以满足日益增长的混合STR图谱分析的需求。本文介绍一套自主研发的混合STR图谱分析系统——SMART(STR Mixture Analysis and Resolution Tools),其能够实现对混合STR图谱的自动化分析,包括拆分混合STR图谱和计算似然比。方法SMART使用的STR峰高的概率模型,考虑了影子峰、降解、基因座特异性扩增效率、峰高变异、插入峰、峰丢失等因素。模型通过比较各个基因型集合拟合当前混合STR图谱的似然值大小来推断最有可能的基因型集合,采用了马尔可夫链蒙特卡洛(MCMC)算法进行求解。结果SMART可应用于不同试剂盒和遗传分析仪产生的数据。经使用ABI-3500XL遗传分析仪和Global Filer试剂盒测试,SMART能够实现对2~5人的混合STR图谱的分析,输出包括混合比例、混合图谱质量、各个贡献者单人的STR基因分型、似然比等多项结果。结论SMART作为一款自主研发的混合STR图谱分析系统,其基本涵盖了国外同类分析系统所具有的功能,技术性能指标达到国际先进水平,能够满足一线法医工作者对于混合STR图谱分析的多样化需求,提高混合STR图谱结果的利用率。

Objective Mixed STR profiles are difficult to analyze for forensic genetics.Most domestic forensic DNA laboratories rely on manual approaches to deconvolve the mixed STR profiles,leaving into a time-consuming and laborious task from which the more or less subjective opinions result in inability to meet the ever-growing demand for analysis of mixed STR profiles.A selfdeveloped analysis system for mixed STR profiles,SMART(STR Mixture Analysis and Resolution Tools),was to introduce here,capable of realizing automatic analysis of mixed STR profiles into their constituents through splitting.Furthermore,SMART is functional of calculating likelihood ratio to unveil whether a suspect is a contributor of the relevant mixed STR profiles.Methods SMART adopts the probabilistic model of STR peak height,taking into account of such factors as stutter,degradation,locus specific amplification efficiency,peak height variation,drop-in and drop-out.With comparison of the likelihood of each assemblage of genotypes to fit the involving mixed STR profile,SMART is to infer the most likely combination of genotypes and uses the Markov Chain Monte Carlo(MCMC)algorithm to support its inference.Results SMART has achieved applications to the data generated from different kits and genetic analyzers.As an example of utilization of ABI-3500 XL genetic analyzer and GlobalFiler kit,SMART has successfully split the mixed STR profiles composed of two to five individuals,outputting multiple resultant items including mixed proportion,mixed profile quality,STR genotyping of each individual contributor and likelihood ratio.Conclusions SMART,the self-developed analysis system for mixed STR profiles,can function to cover similar foreign software systems.Its technical performance indicates that it has reached onto the international advanced level,is eligible to meet the diversified needs of frontline forensic examiners to analyze mixed STR profiles which to have consequently their utilization rate improved.