Inference Of Mixture Proportion In Mixed DNA Analysis

Object:Interpretation of mixture DNA in forensic science is a very challenging problem. It is difficult to separate the various components in the forensic mixture DNA. The results of genotype profiles were usually manifested in the phenomenon of more than two alleles at one locus so that the results of the analysis become complicated. Therefore, the interpretation of forensic DNA mixture needs improving from two aspects including experimental techniques and statistical analysis methods. For the genotype profiles of forensic mixture DNA casework, the first step was to identify the presence of a mixture, then Identify the number of contributors in the mixture, furthermore, considerate all of the possible genotype combinations. Whether the estimation of the mixture proportion was correct or not directly affect the inference results of genotype profiles of each component. Therefore, a key issue in the interpretation of forensic DNA mixture is to infer the mixture proportion. The purpose of this study was to investigate the factors that affecting the inferring of the mixture proportion from two aspects:1)the initial DNA quantitative results with Quantifiler Duo DNA Quantification kit,2) the allele peak information of genotype profiles with the Identifiler kit. Finally, intend to provide valuable guidance and advice to accurately infer the mixture proportion.Methods:1Evaluation of the Hb of single source DNA sample:5ml blood of each180unrelated individuals were collected from Hebei Province Blood Center under informed consent. Whole blood DNA was extracted using QIAampDNA Blood Midi kit. PCR Amplification was performed with AmpF STR Identifiler(?) PCR Amplification Kit on a PTC-200PCR instrument. PCR products were genotyped using ABI3130Genetic Analyzer and analyzed with GeneMappeTM v3.2software.120case samples were selected in our lab in the past year. Finally the heterozygote balance values of the300samples were calculated.2Construction of the simulated mixture DNA models:Samples used in this study included a set of standard DNA abtained from the Promega(9947A and2800M) and two DNA templates extracted with the QIAamp midi kit from blood samples, a male and a female sample named M and F. The four samples were quantified using Quantifiler Duo DNA Quantification kit. According to the quantification results, the concentration of9947A,2800M were diluted to0.5ng/ul, while the M, F to lng/ul. The2800Mand9947A as well as M and F DNA samples, respectively, were mixed at the following mixture ratio:1:13,1:11,1:9,1:7,1:5,1:3,1:1,3:1,5:1,7:1,9:1,11:1,13:1.3Quantification of the mixture DNA samples:The mixed DNA templates were quantified with Quantifiler Duo DNA Quantification kit with three times parallel test. The quantification results were analyzed with AB7500SDS V1.4.3software. Compile statistics of male and total DNA to calculate the mixture proportion to perform relevant statistical analysis.4Genotype mixture DNA samples:Amplification were carried out using AmpF STR Identifiler(?) PCR and PCR products were genotyped with ABI3130Genetic Analyzer(three times), analyzed with GeneMapperTM v3.2software. Information of the peak height of each allele were added up, then according to the formula of the mixture proportion, finally, performed related statistical analysis.5Statistical analysis:Calculate the following parameters:Hb, APH, Mx, Mr, Mx1, D value, D’value, residual, mean residual. Hb is defined as peak height ratio of two different alleles of heterozygous at a locus. Hb1=φ smaller/φ longer, where φ smaller and φ longer represent the peak heights of the smaller and the higher allele at a locus. Average peak height is the half of the sum of two alleles at a locus. Mixture proportion means that the proportion of the one component DNA in the mixture DNA. Mx is the theoretical value. Mx’ is the observed value using real-time PCR. MX1is also the observed value obtained by the genotype profiles. D’value is the difference between the Mx and Mx’while D value is the difference between the Mx and Mx1. Mr is the ratio of each component in the mixture DNA. Residual means the sum of D square. Residual=∑D2. Mean residual’=∑D2/n. n represents the number of examined loci. Data for statistical analysis were performed with Excel and SPSS16.0. The statistical analysis methods incude paired t test,Mann-whiteney U test, Kruakal-wallis test, wilcoxon Signed Ranks test.Results:1Evaluation of the Hb of single source DNA sampleThe genotype profiles of180DNA samples extracted from whole blood were obtained. The following was the distribution of the values of Hb. When the Hb value set as0.4,0.5, and0.6,100%locus were more than0.4,99.9%locus more than0.5,99.5%locus more than0.6.120case samples in the past year were also examined, the values of Hb range was greater than0.4to100%, more than0.5to99.4%, more than0.6to96.7%.300samples of Hb is greater than0.6to98.4%. The median value of D18S51and D2S1338were the lowest,84.6%and85.6%. The rest of the loci are more than86.0%. The median value of the D3S1358and Amelogenin loci were up to90.0%and89.8%, respectively.2Results of Real-time PCR using Quantifiler Duo Quantification kitThere was no significant difference between theoretical values and observed values in mixture proportion of2800M-9947A DNA mixture and significant difference was detected in the M-F DNA mixture. The D’values of Mr<1group and Mr>1group were both less than0.11,while significant difference could be observed between the two groups. The D’values of Mr>1Group were greater than Mr<1group.With the proportion of male DNA component in the mixture DNA increasing, D’values also increased. In the M-F DNA mixture, when the mixture ratio reached to11:1and13:1, the results of real-time PCR showed that the amount of male DNA were more than the total DNA. So the values of mixture proportion could not be calculated, indicating that if the mixture proportion of male DNA is more than91%, it may not obtain accurate mixture proportion with Quantifiler Duo DNA quantification kit.3The factors that affecting the inferring of the mixture proportion according to the genotype profiles with Identifiler kit3.1The effects of the initial mixtureon inferring the MX1Compare the D values of different mixture ratio. There was no significant different between the6mixture ratio in Mr<1group and the6mixture ratio in the Mr>1group. The difference in Mr>1group are greater than the Mr<1group in2800M-9947A DNA mixture,but there was no significant difference between the Mr>1group and Mr<1group in M-F DNA mixture.In the mixture DNA samples of the2800M-9947A, For D7S820,the D values were high, among them7D values were higher than0.25,4D value were greater than0.35. The rest15loci were below0.35. Dropout were discovered in D21S11loci at1:13and D16S539loci at13:1mixture ratio.95.1%of the D value is less than0.10. About the M-F mixture DNA samples, all the D values were below0.35. There was no dropout phenomenon and96.2%of the D value is less than0.10.3.2The effects of different locuson inferring the Mx’For the2800M-9947A and M-F mixture DNA samples, D value and residual of the15STR loci and Amelogenin were calculated. Residual values were examined with Kruskal-wallis test, D values of different loci for inferring the mixture proportion were different. The Mx and Mx1of16loci were compared with paired t test and wilcoxon Signed Ranks test.The results indicated as following:For CSF1PO, D5S818, D16S539, D18S51, FGA and vWA, There was no significant statistical difference between Mx and Mx1and For the rest loci, there was significant statistical difference. For the M-F mixture DNA, D2S1338, D7S820,D16S539, D18S51, D19S433,D21S11, TPOx, vWA, There was no significant statistical difference between Mx and Mx1, For the2800M-9947A and M-F mixture DNA samples, vWA, D18S51, D16S539has no significant difference. Amelogenin, D3S1358, D13S317, Th01, D8S1179. The residual values of vWA, D18S51, D16S539were more lower, and Amelogenin, D3S1358, D13S317, Th01,D8S1179were higher. 3.3The effects of alleles sharing on the Mx1In the2800M-9947A mixture DNA, the mean residual value of the loci with alleles sharing was0.182, the loci without alleles sharing was0.022. The results indicated that the mean residual value of loci without alleles sharing was lower than the loci with alleles sharing. We further compare the mean residual of different genotype combinations with no alleles sharing, the mean residual of4alleles was0.025,3alleles0.023,2alleles0.012. In short, For the loci without alleles, the mean residual of2alleles (BB:AA) was the most lowest.In the M-F mixture DNA, the mean residual value of the loci with alleles sharing was0.048, without alleles sharing was0.012. In the form of loci without alleles sharing,4alleles was0.015,3alleles0.010,2alleles0.008.It is more accurate to inferring the mixture proportion for no alleles sharing than alleles sharing. In the three forms of non-allele sharing, The2alleles for inferring the mixture proportion was the most accurate. So the loci with alleles sharing or without alleles sharing has an effects on the interring the mixture proportion.Conclusions:1When analyzing mixture DNA, we recommended firstly quantify the initial concentration of the mixed DNA and obtain the mixture proportion of the two samples using Quantifiler Duo DNA Quantification kit or other kits. However, the mixture ratio obtained by the kits is more accurate when female samples are the major component in the mixed DNA. If the male DNA is the major component, the mixture ratio obtained by the kits is not so accurate. When mixture proportion of the male DNA is more than91%, the data of mixture may not be obtained.2When Inferring the mixiture proportion based on genotype profiles, locus without allele sharing is more accurate than locus with alleles sharing. Therefore, amelogenin is not the best choice to estimate the mixture proportion. We recommended it is firstly to select the locus without alleles sharing to estimate the mixture proportion.