从近亲干扰角度对亲缘鉴定标准的研究

发布时间：2018-06-27 18:22

本文选题：法医物证学 + 亲缘鉴定　；参考：《山西医科大学》2011年硕士论文

【摘要】：目的:从其他亲缘关系对常规亲子鉴定干扰的角度,研究亲缘鉴定的判定标准,在需要检验基因座的总数量、排除概率与似然比率的校正等方面探讨在考虑近亲干扰时亲子鉴定的评估策略。方法:本课题研究内容共分五部分。 1、调查一个家系,共四代25人。获得常用15个基因座(CODIS+D2S1338,D19S433)的检测分型结果,模拟叔侄关系替代父子关系构成三联体鉴定,统计易出现误判(矛盾基因座数≤2)的案件比例。 2、选用亲属排除概率公式PER=(k_0+k_1)×PE与叔侄(祖孙、半同胞)关系的k系数校正单基因座排除概率;计算常用19个STR基因座(CODIS+D2S1338,D19S433,Penta D,Penta E,D12S391,D6S1043)的平均排除概率和累积排除概率,并与二项式分布定律公式联合应用,推算考虑近亲干扰时亲缘鉴定理论上至少应检测的STR基因座数量。 3、在北京地区人群中调查21个STR基因座(AGCU21+1系统)的群体遗传资料,并与常用19个STR基因座联合应用于常规亲子鉴定中遭遇近亲干扰的典型案例。 4、通过似然比率校正公式LR=k_0+(1-k_0)PI和典型父权指数(PI_(typical))推导在考虑其他亲缘关系干扰时累积似然比率的快速估算法,并用此方法在2389例亲子鉴定中评估考虑近亲干扰时的风险。 5、根据113个全同胞对和113个无关个体对的19个STR基因座分型结果,统计全同胞组和无关个体组的共有等位基因数,使用SPSS软件对两组数据进行判别分析。结果:通过上述研究,主要得到以下结果。 1、本次研究中叔侄假冒父子关系参加亲缘鉴定时的检测结果表明:常规15个基因座时容易误判的案件比例在三联体鉴定中为17%;在二联体鉴定中为37.5%。 2、考虑叔侄、祖孙、半同胞三种亲缘关系干扰影响,校正后的单基因座PE值是校正前的0.75倍。为使累积似然比率达到1×10~4,在三联体鉴定中,应至少检验15个STR基因座(单基因座排除概率0.481);当检出1个矛盾基因座时,检测基因座总数应增加至24个;检出2个矛盾基因座时,检出总数应增加至35个;检出3个矛盾基因座时,检出总数应增加至43个。在二联体鉴定中,应至少检验21个STR基因座(单基因座排除概率0.358);当检出1个矛盾基因座时,检测基因座总数应增加至34个;检出2个矛盾基因座时,检出总数应增加至48个。 3、本次群调新增20个非CODIS基因座(D19S433重复)在北京地区人群中的遗传资料,可与常用19个基因座联合应用。在遭遇近亲干扰的亲缘鉴定案例中,当检测基因座总数达到结果2中要求时,均避免了将矛盾基因误判为遗传变异。 4、在亲缘鉴定中用累积似然比率(CPI)评估检测结果时,如果考虑到近亲干扰的可能,可以通过CPI×0.75~(n-m)快速估算校正后的累积似然比率(n为检测基因座总数,m为矛盾基因座个数)。当出现基因座符合遗传规律但PI值1的情况时(母子表型相同时可能发生),该估算方法误差较大。 5、在检验总数为19个STR基因座时,利用共有等位基因数(S)建立的判别函数与判别法则(L同胞=3.3408×S-39.8808;L无关=1.6685×S-10.4679;L(S)=max[L同胞(S)+ L无关(S)])可用于评估两个体间是否为同胞关系,评估指标为后验概率。在S14和S21时,后验概率达到99%以上,它们分别占到无关个体组的89.3%和全同胞组的83.2%;在S=16或17时可能出现判别错误,错误率约0.9%。结论:在亲缘鉴定中,当孩子的某个非父近亲代替孩子真正亲代参与鉴定时,只检测常用15个STR基因座将存在误判风险。当检测基因座总数达到本研究推导出的数量要求时,可以避免因近亲干扰而造成错误。在评估STR检测结果对父权假设的支持程度时,如果考虑近亲干扰,可以通过本文提供的方法快速校正累积似然比率,使能够大致把握检测结果的证据强度。在涉及全同胞的鉴定中,可以通过共有等位基因数进行辅助判别。本文推导的方法可以在一定程度上避免其他亲缘关系对常规亲子鉴定的干扰,为进一步研究亲子鉴定的标准提供了依据。
[Abstract]:Objective: To study the evaluation criteria of paternity identification from the angle of interference of other relatives to conventional paternity identification, and to study the evaluation criteria of paternity identification, and to discuss the evaluation strategy of paternity testing in consideration of the interference of close relatives in the aspects of the total number of loci, the correction of the probability of exclusion and the likelihood ratio.
Methods: the contents of this study were divided into five parts.
1, an investigation of a family, a total of four generations, 25 people, obtained the results of the detection and classification of common 15 CODIS+D2S1338 (D19S433), and simulated the relationship between nephew and nephew instead of father and son to form three body identification, and the proportion of cases of misjudgment (the number of contradictory loci less than 2) was easily found.
2, using the relative exclusion probability formula PER= (k_0+k_1) x PE and the relationship between the uncle and nephew (Zu Sun, half compatriots) to correct the exclusion probability of the single locus, and calculate the average exclusion probability and the cumulative exclusion probability of the 19 commonly used STR loci (CODIS+D2S1338, D19S433, Penta D, Penta E, D12S391), and combined with the binomial distribution law formula. The number of STR loci that should be detected in phylogenetic analysis should be considered when considering close relatives interference.
3, the population genetic data of 21 STR loci (AGCU21+1 system) were investigated in the Beijing population, and the typical cases of close relatives interference with the common 19 STR loci were used in the routine paternity test.
4, by using the likelihood ratio correction formula LR=k_0+ (1-k_0) PI and the typical paternity index (PI_ (typical)), the fast estimation method of cumulative likelihood ratio when considering the interference of other kinship relationships is derived, and the risk of close relatives interference is evaluated with this method in 2389 paternity tests.
5, according to the results of the 19 STR genotyping results from 113 full compatriots and 113 unrelated individuals, the total number of alleles in the whole compatriot group and the unrelated individual group was counted, and the data of the two groups were discriminate with the SPSS software.
Results: through the above research, the following results were obtained.
1, in this study, the test results of the relationship between the uncle and nephew of the uncle and nephew in the study showed that the proportion of the cases that were easily misjudged when the conventional 15 loci were misjudged was 17% in the three conjoint identification, and in the two conjoined identification was 37.5%.
2, considering the influence of nephew, grandson, grandson and half sibling, the PE value of the corrected monogenic pedestal is 0.75 times that before correction. In order to make the cumulative likelihood ratio reach 1 x 10~4, at least 15 STR loci should be tested at least 15 STR loci (single locus exclusion probability 0.481); when 1 contradictory loci are detected, the total number of detected loci should be increased. Add to 24; the total number of detectable 2 loci should be increased to 35; when 3 contradictory loci were detected, the total number should be increased to 43. At least 21 STR loci (Dan Jiyin exclusion probability 0.358) should be tested in two conjoined identification. When 1 contradictory base loci were detected, the total number of detected loci should be increased to 34; 2 detection should be detected. When the locus of the conflicting loci, the total number of detection should be increased to 48.
3, the genetic data in the population of 20 non CODIS loci (D19S433 repeats) in Beijing region can be combined with 19 common loci. In the case of consanguineous family identification with close relatives, when the total number of loci was detected in the result 2, the conflicting genes were misjudged to be genetic variation.
4, when using the cumulative likelihood ratio (CPI) to assess the test results, the cumulative likelihood rate after correction can be quickly estimated by CPI x 0.75~ (n-m) (n is the number of detected loci, M is the number of contradictory genes). It may occur at the same time, and the estimation method has a large error.
5, when the total number of 19 STR loci is tested, a discriminant function and a discriminant rule (L compatriot =3.3408 x S-39.8808; L independent =1.6685 x S-10.4679; L (S) =max[L compatriot (S) + unrelated) can be used to evaluate whether the two individuals are siblings, and the evaluation index is a posterior probability. Up to 99%, they accounted for 83.2% of the unrelated individuals and 83.2% of the full sib group, and there might be a discriminant error at S=16 or 17, with an error rate of about 0.9%..
Conclusion: in the paternity test, only 15 common STR loci will be found to be misjudged when a child's non parent relatives substitute for a child to participate in the identification. When the total number of loci reaches the number requirements derived from this study, the error caused by close relatives can be avoided. The paternity hypothesis of the STR test results is evaluated. In the support degree, if close relatives are considered, the cumulative likelihood ratio can be quickly corrected by the method provided in this paper, so that the evidence strength of the results can be roughly grasped. In the identification of all compatriots, the number of common alleles can be used to discriminate. The method derived in this paper can avoid others to a certain extent. The influence of genetic relationship on routine paternity testing provides a basis for further study of paternity testing criteria.
【学位授予单位】：山西医科大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：D919

【参考文献】