Applications of 1993 single nucleotide polymorphism loci in forensic pairwise kinship identifications and inferences

Kinship testing plays critical roles in criminal investigations, missing person searches, civil disputes, as well as identifying disaster victims. The existing commonly used short tandem repeat (STR) loci have limited effectiveness in the identification of second-degree and more distant kinships. In this study, a total of 1993 SNP loci of 119 Chinese Han individuals from eight families were sequenced on the MGISEQ-2000RS platform. The system powers of this panel for kinship identifications were evaluated based on both the likelihood ratio (LR) and identical by state (IBS) methods. The results indicated that this panel could be used as an effective tool to kinship analyses including paternity testing, full sibling testing, second-degree kinships, and first cousin kinship analyses. Both the LR and IBS methods could be applied in distinguishing first-degree and second-degree pairs from unrelated individuals. Based on the 1993 SNP loci, LR>1000 and LR<0.001 are recommended as the thresholds of identifying first-cousin kinships from unrelated individuals, and the system power of such thresholds was 0.9470. Besides, kinship coefficients for different kinship pairs were estimated and then were used to predict the kinships for pairwise individuals. This panel performs an effective kinship inference power for the predictions of first-degree, second-degree kinships and unrelated individual pairs, while presenting low sensitivity in the prediction of first-cousin kinships.     

Efficacy of extended kinship analyses utilizing commercial STR kit in establishing personal identification

Unprecedented fidelity and specificity have afforded DNA testing its long reigning status as the gold standard for establishing personal identification. While the method itself is flawless, forensic experts have undoubtedly stumbled across challenging cases in which no reference samples for an unknown person (UP) are available for comparison. In such cases, experts often must resort to an assortment of kinship analyses-primarily those involving alleged parents or children of a UP-to establish personal identification. The present study derives likelihood ratio (LR) distributions from an extensive series of kinship simulations and places actual data, obtained from 120 cases in which personal identification of a UP was established via kinship analyses, to a comprehensive comparison in order to evaluate the efficacy of kinship assessments in establishing personal identification. A commercially available AmpFlSTR Identifiler kit was used to obtain DNA profiles. UP DNAs were extracted and isolated from fingernail (n=87), cardiac blood (24), carpal bone (7) and tooth (2). Buccal cells were procured from alleged kin (AK) for subsequent kinship analyses. In 72 cases 1-3 alleged children were available for comparison; in 46 cases, one or both alleged parents were available; and in the final 2 cases (involving a pair of bodies discovered together in a dwelling), their alleged children were typed for comparison. For each case a LR was calculated based on the DNA typing results. Interestingly, we found that the median LR observed in the actual cases virtually mirrored those of the simulations. With exception to 2 cases in which a silent allele was observed at D19S433, biological relatives showed a LR greater than 100 and in these cases, kinship between the UP and AK were further supported by additional forms of evidence. We show here that in the vast majority of identification cases where direct reference samples are unavailable for a UP, kinship analyses referring to alleged parents/children and using 15 standard loci is more than capable of establishing the identification of a UP. However, discretion should be advised for silent alleles which-albeit rare-are known to occur at loci such as D19S433, along with other mutations which could render a deceivingly reduced LR.     

Likelihood ratio and posterior odds in forensic genetics: Two sides of the same coin

It has become widely accepted in forensics that, owing to a lack of sensible priors, the evidential value of matching DNA profiles in trace donor identification or kinship analysis is most sensibly communicated in the form of a likelihood ratio (LR). This restraint does not abate the fact that the posterior odds (PO) would be the preferred basis for returning a verdict. A completely different situation holds for Forensic DNA Phenotyping (FDP), which is aimed at predicting externally visible characteristics (EVCs) of a trace donor from DNA left behind at the crime scene. FDP is intended to provide leads to the police investigation helping them to find unknown trace donors that are unidentifiable by DNA profiling. The statistical models underlying FDP typically yield posterior odds (PO) for an individual possessing a certain EVC. This apparent discrepancy has led to confusion as to when LR or PO is the appropriate outcome of forensic DNA analysis to be communicated to the investigating authorities. We thus set out to clarify the distinction between LR and PO in the context of forensic DNA profiling and FDP from a statistical point of view. In so doing, we also addressed the influence of population affiliation on LR and PO. In contrast to the well-known population dependency of the LR in DNA profiling, the PO as obtained in FDP may be widely population-independent. The actual degree of independence, however, is a matter of (i) how much of the causality of the respective EVC is captured by the genetic markers used for FDP and (ii) by the extent to which non-genetic such as environmental causal factors of the same EVC are distributed equally throughout populations. The fact that an LR should be communicated in cases of DNA profiling whereas the PO are suitable for FDP does not conflict with theory, but rather reflects the immanent differences between these two forensic applications of DNA information.     

Enhancing accurate data collection in mass fatality kinship identifications: Lessons learned from Hurricane Katrina

A mass fatality DNA identification effort is a complex process in which direct matching and kinship analysis is used for identifying human remains. Kinship DNA identification is an important tool in the identification process in which victim's DNA profiles are compared to the profiles of “known” biologically related reference samples. Experience from the 9/11 World Trade Center DNA identification efforts showed that forms used to record biological relationships are important and that inaccurately documented information may hamper the kinship analysis and DNA identification process. In the identification efforts following Hurricane Katrina, a Family and/or Donor Reference Collection (FDRC) form was used as a means to document the reported relationship between the reference DNA donor and the purported missing individual. This FDRC form was developed based upon lessons learned from 9/11 and the Tsunami identification efforts. This paper analyses the effectiveness of the FDRC form used in the Hurricane Katrina kinship DNA identification efforts and proposes an improved sample collection form for kinship and other donor reference samples. The data presented can be used to enhance the accuracy of the data collection process through an improved sample collection form, streamlining the DNA kinship identification process and decreasing the burden on valuable resources.     

ProbRank: An efficient DNA database search method for complex mixtures per a quantitative likelihood ratio model

Searching a DNA Database with a DNA profile from an evidentiary trace can provide investigative leads in a forensic case. Various searching approaches exist such as conventional methods based on matching alleles or more advanced methods computing likelihood ratios (LR) while considering drop-in and drop-out. Here we examine the potential of using a quantitative LR model (EuroForMix model incorporated in ProbRank method) that takes peak heights into account in comparison to a qualitative LR model (LRmix model implemented in SmartRank method). Both methods present DNA database candidates in order of decreasing LR. Especially regarding minor contributors in complex mixtures, the method using the quantitative model outperforms the method using the qualitative model in terms of sensitivity and specificity as more true donors and less adventitious matches are retrieved. ProbRank is to be implemented in DNAStatistX and is sufficiently fast for daily use.     

Accurate assessment of the weight of evidence for DNA mixtures by integrating the likelihood ratio

Several methods exist for weight of evidence calculations on DNA mixtures. Especially if dropout is a possibility, it may be difficult to estimate mixture specific parameters needed for the evaluation. For semi-continuous models, the LR for a person to have contributed to a mixture depends on the specified number of contributors and the probability of dropout for each. We show here that, for the semi-continuous model that we consider, the weight of evidence can be accurately obtained by applying the standard statistical technique of integrating the likelihood ratio against the parameter likelihoods obtained from the mixture data. This method takes into account all likelihood ratios belonging to every choice of parameters, but LR's belonging to parameters that provide a better explanation to the mixture data put in more weight into the final result. We therefore avoid having to estimate the number of contributors or their probabilities of dropout, and let the whole evaluation depend on the mixture data and the allele frequencies, which is a practical advantage as well as a gain in objectivity. Using simulated mixtures, we compare the LR obtained in this way with the best informed LR, i.e., the LR using the parameters that were used to generate the data, and show that results obtained by integration of the LR approximate closely these ideal values. We investigate both contributors and non-contributors for mixtures with various numbers of contributors. For contributors we always obtain a result close to the best informed LR whereas non-contributors are excluded more strongly if a smaller dropout probability is imposed for them. The results therefore naturally lead us to reconsider what we mean by a contributor, or by the number of contributors.     

Validation of SmartRank: A likelihood ratio software for searching national DNA databases with complex DNA profiles

Searching a national DNA database with complex and incomplete profiles usually yields very large numbers of possible matches that can present many candidate suspects to be further investigated by the forensic scientist and/or police. Current practice in most forensic laboratories consists of ordering these ‘hits’ based on the number of matching alleles with the searched profile. Thus, candidate profiles that share the same number of matching alleles are not differentiated and due to the lack of other ranking criteria for the candidate list it may be difficult to discern a true match from the false positives or notice that all candidates are in fact false positives. SmartRank was developed to put forward only relevant candidates and rank them accordingly. The SmartRank software computes a likelihood ratio (LR) for the searched profile and each profile in the DNA database and ranks database entries above a defined LR threshold according to the calculated LR. In this study, we examined for mixed DNA profiles of variable complexity whether the true donors are retrieved, what the number of false positives above an LR threshold is and the ranking position of the true donors. Using 343 mixed DNA profiles over 750 SmartRank searches were performed. In addition, the performance of SmartRank and CODIS were compared regarding DNA database searches and SmartRank was found complementary to CODIS. We also describe the applicable domain of SmartRank and provide guidelines. The SmartRank software is open-source and freely available. Using the best practice guidelines, SmartRank enables obtaining investigative leads in criminal cases lacking a suspect.     

Effects of DNA degradation and genotype imputation on high-density SNP microarray in pairwise kinship analysis

High-density single nucleotide polymorphisms (SNPs) can detect distant relatives even in the context of pairwise kinship analysis. Although DNA microarrays conveniently generate genome-wide SNP data, they require large quantities of high-quality DNA. Genotyping data obtained from low-quantity and low-quality samples are likely unreliable owing to the incidence of no-called or mistyped SNPs. In this study, we examined the effects of insufficient sample densities and sample degradation on the efficacy of kinship analysis. While low DNA amounts had a minor effect, DNA degradation led to a significant increase in no-call rates and error rates. Posterior probabilities of kinship determination, calculated using the index of chromosomal sharing, were markedly lower in proportion to the no-call rates and error rates. We also investigated the effect of genotype imputation to complement the no-called genome data utilizing SNPs reference panels. We found that the posterior probability of the relative-assumed person increased with genotype complementation in case of mild degradation, even with mistyped genotypes. Therefore, DNA microarray with imputation is a promising method for analyzing forensic DNA samples taken from situations where DNA quantity and quality may be compromised, such as disaster victim identification using pairwise kinship analysis.     

Assessing the utility of quantitative and qualitative metrics in the DNA quantification process of skeletal remains for autosomal and Y-chromosome STR amplification purposes

In historical cases, ancient DNA investigations and missing persons identification, teeth or bone samples are often the only and almost always the best biological material available for DNA typing. On the other hand, DNA obtained from bone material may be characterized by a high degradation index (DI) or its low content, or DNA tests cannot be repeated due to bone piece size limitation. That is often the effect of the environment in which the material was placed and the time during which exposure to unfavorable environmental factors took place. Therefore, it is very important to use appropriate procedures related to STR analysis. For our study, we selected 80 challenging bone samples. The amount of DNA was compared in qPCR using Quantifiler™ Trio DNA Quantification Kit and Investigator® Quantiplex® Pro RGQ. All qPCR results were confirmed by PCR-CE. The results of DNA concentrations and the assigned degradation index (DI) differed significantly within analyzed samples (~10%). Additionally, the Y-chromosome DI also differed from the autosomal DI in the samples. The difference in degradation indexes could explain the lower Y-chromosome amplification success rate compared to autosomal e.g. during human identification process. The results indicate that performing two DNA quantifications with the use of two different kits (primers sets) allows for a much more precise evaluation of the DNA quality and quantity in the isolate. We suggest that at least one of two suggested DNA concentration measurements should be based on an additional determination of the Y chromosome degradation index. Altogether, it allows for rational isolate management, especially when the volume is limited and the sample is unique.     

Exploring the DNA mixture deconvolution through simulation

ABSTRACT

If an unambiguous single-source DNA profile is obtained from a crime scene, then a potential person of interest can either match or not match the crime scene profile and the likelihood ratio for the single matching genotype can be easily computed. Mixed DNA profiles on the other hand are typically ambiguous and a vast number of different likelihood ratios can be obtained depending on the genotype of a potential person of interest that is compared with the mixture later. In the absence of a person of interest it can be unclear how suitable the profile is for discriminating between donors and non-donors. We introduce a simulation method to explore the range of likelihood ratios that is expected to be obtained when a non-donor or a true donor is compared with the mixed DNA profile. Sampling is conditional on the mixture deconvolution obtained using probabilistic genotyping. These simulations help to decide whether or not a (mixed) profile is suitable for comparison to a person of interest. Moreover, the methods can be used to determine whether a profile is suitable for upload to a database and whether or not potential rework could be advised.     

Forensic botany: species identification of botanical trace evidence using a multigene barcoding approach

Forensic botany can provide significant supporting evidence during criminal investigations. However, it is still an underutilized field of investigation with its most common application limited to identifying specific as well as suspected illegal plants. The ubiquitous presence of plant species can be useful in forensics, but the absence of an accurate identification system remains the major obstacle to the present inability to routinely and correctly identify trace botanical evidence. Many plant materials cannot be identified and differentiated to the species level by traditional morphological characteristics when botanical specimens are degraded and lack physical features. By taking advantage of a universal barcode system, DNA sequencing, and other biomolecular techniques used routinely in forensic investigations, two chloroplast DNA regions were evaluated for their use as “barcoding” markers for plant identification in the field of forensics. We therefore investigated the forensic use of two non-coding plastid regions, psbA-trnH and trnL-trnF, to create a multimarker system for species identification that could be useful throughout the plant kingdom. The sequences from 63 plants belonging to our local flora were submitted and registered on the GenBank database. Sequence comparison to set up the level of identification (species, genus, or family) through Blast algorithms allowed us to assess the suitability of this method. The results confirmed the effectiveness of our botanic universal multimarker assay in forensic investigations.     

The comparison of DNA mixture profiles with multiple persons of interest

When we compare a DNA mixture profile to a single person of interest, there are often just two competing explanations considered, and the comparison of how likely these are to lead to the observed mixture is summarized by a likelihood ratio. However, in more complex cases this does not suffice, e.g., when there are multiple persons of interest. One can then compute several likelihood ratios, corresponding to several pairs of hypotheses, and subsequently decide which one(s) to report. This may lead to the computation of a rather large number of such likelihood ratios. In this article we advocate a systematic approach that starts by describing all relevant hypotheses. For each hypothesis, we then compute its likelihood (i.e., the probability to see the genetic data if the hypothesis is true). Based on the likelihoods of all considered hypotheses, one can then make a summary of the findings to report. This may be on the level of the considered hypotheses and/or with likelihood ratios per person of interest. We illustrate with several examples how this approach assists interpretation. The likelihoods summarize how the trace can help to distinguish between the considered hypotheses, in the sense that they transform the prior odds on them into posterior odds, without having to assign prior probabilities on the hypotheses for the calculation of the likelihoods themselves. On the other hand likelihood ratios (LR’s) for individual PoI’s cannot be obtained without these priors. In many cases these LR’s will be quite insensitive to the choice of prior probabilities but in other cases they will be; we give examples of both.We argue that the table of likelihoods of the considered hypotheses is a more natural analog of the LR provided in the simple case with one PoI and two considered hypotheses, compared to the computation of a LR per PoI. We end with a discussion of the choice of prior probabilities, of the existing recommendations for this situation, and on reporting.     

GHEP-ISFG collaborative simulated exercise for DVI/MPI: Lessons learned about large-scale profile database comparisons

The GHEP-ISFG Working Group has recognized the importance of assisting DNA laboratories to gain expertise in handling DVI or missing persons identification (MPI) projects which involve the need for large-scale genetic profile comparisons. Eleven laboratories participated in a DNA matching exercise to identify victims from a hypothetical conflict with 193 missing persons. The post mortem database was comprised of 87 skeletal remain profiles from a secondary mass grave displaying a minimal number of 58 individuals with evidence of commingling. The reference database was represented by 286 family reference profiles with diverse pedigrees. The goal of the exercise was to correctly discover re-associations and family matches. The results of direct matching for commingled remains re-associations were correct and fully concordant among all laboratories. However, the kinship analysis for missing persons identifications showed variable results among the participants. There was a group of laboratories with correct, concordant results but nearly half of the others showed discrepant results exhibiting likelihood ratio differences of several degrees of magnitude in some cases. Three main errors were detected: (a) some laboratories did not use the complete reference family genetic data to report the match with the remains, (b) the identity and/or non-identity hypotheses were sometimes wrongly expressed in the likelihood ratio calculations, and (c) many laboratories did not properly evaluate the prior odds for the event. The results suggest that large-scale profile comparisons for DVI or MPI is a challenge for forensic genetics laboratories and the statistical treatment of DNA matching and the Bayesian framework should be better standardized among laboratories.     

Forensic identification of an individual in complex DNA mixtures

The identification of a suspect in a complex DNA mixture typed with standard short tandem repeat (STR) kits has proved difficult. In the current study we present the theoretical framework of a method aimed to resolve this problem in forensic cases. The method suggests genotyping a specially designed panel of 1000–3000 single nucleotide polymorphisms (SNPs), each with a relatively low (<0.1) minor allele frequency (MAF). The rationale of this method is that any individual will carry a specific set of dozens of rare alleles and the complex DNA mixture will carry this particular set only if the one individual is represented in the DNA mixture. The efficiency of the method is evaluated by estimating the probability that a random man will not be excluded (RMNE) from the mixture. When this probability, P(RMNE), is low, one can conclude that the suspect's DNA is present in the DNA mixture. Essentially, a P(RMNE) < 10⁻⁹ is considered as proof, whereas a P(RMNE) < 10⁻⁶ is considered strong evidence. For completeness, we also analyzed the method using the likelihood ratio (LR) approach. We have analyzed the method for a variety of conditions and found that generally the method will provide highly significant results even for complex mixtures combining up to 10 individuals. The method performs well even when close relatives (one or two brothers) are present in the complex DNA mixture and when contributors or suspects come from different populations. We have also found that the method can accurately identify the number of contributors to the mixture, something that in some instances has significant forensic value on its own.     

Generating STR profile from "Touch DNA"

As forensic DNA technology has become a common tool in criminal investigations, scientists have attempted to obtain DNA evidence from what were once considered unlikely sources. “Touch DNA” refers to the DNA that is left behind from skin cells when a person touches or comes into contact with an item. This present study shows, DNA profiling of touched evidence materials is reported employing a combination of LCN typing and miniSTRs. The technology is highly valuable for increasing the scope of DNA profiling to large number touched evidence materials.     

Bringing colour back after 70 years: Predicting eye and hair colour from skeletal remains of World War II victims using the HIrisPlex system

Retrieving information about externally visible characteristics from DNA can provide investigative leads to find unknown perpetrators, and can also help in disaster victim and other missing person identification cases. Aiming for the application to both types of forensic casework, we previously developed and forensically validated the HIrisPlex test system enabling parallel DNA prediction of eye and hair colour. Although a recent proof-of-principle study demonstrated the general suitability of the HIrisPlex system for successfully analysing DNA from bones and teeth of various storage times and conditions, practical case applications to human remains are scarce. In this study, we applied the HIrisPlex system to 49 DNA samples obtained from bones or teeth of World War II victims excavated at six sites, mostly mass graves, in Slovenia. PCR-based DNA quantification ranged from 4 pg/μl to 313 pg/μl and on an average was 41 pg/μl across all samples. All 49 samples generated complete HIrisPlex profiles with the exception of one MC1R DNA marker (N29insA) missing in 83.7% of the samples. In 44 of the 49 samples (89.8%) complete 15-loci autosomal STR (plus amelogenin) profiles were obtained. Of 5 pairs of skeletal remains for which STR profiling suggested an origin in the same individuals, respectively, 4 showed the same HIrisPlex profiles and predicted eye and hair colours, respectively, while discrepancies in one pair (sample 26 and 43) are likely to be explained by DNA quantity and quality issues observed in sample 43. Sample 43 had the lowest DNA concentration of only 4 pg/μl, producing least reliable STR results and could be misleading in concluding that samples 43 and 26 originate from the same individual. The HIrisPlex-predicted eye and hair colours from two skeletal samples, suggested to derive from two brothers via STR profiling together with a living sister, were confirmed by the living sister's report. Overall, we demonstrate that after more than 70 years, HIrisPlex-based eye and hair colour prediction from skeletal remains is feasible with high success rate. Our results further encourage the use of the HIrisPlex system in missing person/disaster victim identification to aid the identification process in cases where ante-mortem samples or putative relatives are not directly available, and DNA predicted eye and hair colour information provides leads for locating them, allowing STRbased individual identification.     

Impact of SNP microarray analysis of compromised DNA on kinship classification success in the context of investigative genetic genealogy

Single nucleotide polymorphism (SNP) data generated with microarray technologies have been used to solve murder cases via investigative leads obtained from identifying relatives of the unknown perpetrator included in accessible genomic databases, an approach referred to as investigative genetic genealogy (IGG). However, SNP microarrays were developed for relatively high input DNA quantity and quality, while DNA typically obtainable from crime scene stains is of low DNA quantity and quality, and SNP microarray data obtained from compromised DNA are largely missing. By applying the Illumina Global Screening Array (GSA) to 264 DNA samples with systematically altered quantity and quality, we empirically tested the impact of SNP microarray analysis of compromised DNA on kinship classification success, as relevant in IGG. Reference data from manufacturer-recommended input DNA quality and quantity were used to estimate genotype accuracy in the compromised DNA samples and for simulating data of different degree relatives. Although stepwise decrease of input DNA amount from 200 ng to 6.25 pg led to decreased SNP call rates and increased genotyping errors, kinship classification success did not decrease down to 250 pg for siblings and 1st cousins, 1 ng for 2nd cousins, while at 25 pg and below kinship classification success was zero. Stepwise decrease of input DNA quality via increased DNA fragmentation resulted in the decrease of genotyping accuracy as well as kinship classification success, which went down to zero at the average DNA fragment size of 150 base pairs. Combining decreased DNA quantity and quality in mock casework and skeletal samples further highlighted possibilities and limitations. Overall, GSA analysis achieved maximal kinship classification success from 800 to 200 times lower input DNA quantities than manufacturer-recommended, although DNA quality plays a key role too, while compromised DNA produced false negative kinship classifications rather than false positive ones.     

A new approach for forensic analysis of saliva-containing body fluid mixtures based on SNPs and methylation patterns of nearby CpGs

Saliva samples obtained from crime scenes often contain body fluids from other people, which makes it difficult to not only interpret the obtained DNA profiles, but also interpret saliva identification test results. α-amylase activity, an indicator of most saliva identification methods, can be slightly detected in other body fluids. This study aimed to overcome these difficulties. Here, we identified 13 saliva-specific methylated regions and five saliva-specific unmethylated regions neighboring common single nucleotide polymorphisms (SNPs) by array-based genome-wide methylation analysis of pooled saliva, blood, semen, or vaginal swab samples. Bisulfite sequencing by massively parallel sequencing (MPS) technology was then performed using individual body fluid samples to evaluate the saliva-specificity of each CpG of the three regions selected from the identified candidates. Although no single CpG demonstrated complete saliva-specificity, we found that the reads that were simultaneously (un)methylated at the selected neighboring two to three CpGs of each region were highly specific for saliva DNA. Based on these findings, we then designed MPS-based bisulfite sequencing assays for each region to analyze the selected CpGs and SNP(s) on the same read. These assays could identify the saliva of a target person from body fluid mixtures of known contributors (individual-specific saliva identification) by calculating the ratios of simultaneous (un)methylation at the selected CpGs within the reads containing SNP alleles unique to the target person. Moreover, these assays could indicate the SNP types of saliva DNA (saliva-specific genotyping) from body fluid mixtures by analyzing the alleles of the reads simultaneously (un)methylated at the selected CpGs, while careful attention should be paid to interpret the results of heterologous genotypes. Although further regions should be identified, especially for saliva-specific individual identification, the CpG-SNP approach may be an effective method to interpret the complicated results obtained from saliva-containing body fluid mixtures.     

Evidentiary evaluation of single cells renders highly informative forensic comparisons across multifarious admixtures

The consistency between DNA evidence and person(s) of interest (PoI) is summarized by a likelihood ratio (LR): the probability of the data given the PoI contributed divided by the probability given they did not. It is often the case that there are several PoI who may have individually or jointly contributed to the stain. If there is more than one PoI, or the number of contributors (NoC) cannot easily be determined, then several sets of hypotheses are needed, requiring significant resources to complete the interpretation.Recent technological developments in laboratory systems offer a way forward, by enabling production of single cell data. Though single-cell data may be procured by next generation sequencing or capillary electrophoresis workflows, in this work we focus our attention on assessing the consistency between PoIs and a collection of single cell electropherograms (scEPGs) from diploid cells – i.e., leukocytes and epithelial cells. Specifically, we introduce a framework that: I) clusters scEPGs into collections, each originating from one genetic source; II) for each PoI, determines a LR for each cluster of scEPGs; and III) by averaging the likelihood ratios for each PoI across all clusters provides a whole-sample weight of evidence summary. By using Model Based Clustering (MBC) in step I) and an algorithm, named EESCIt for Evidentiary Evaluation of Single Cells, that computes single-cell LRs in step II), we show that 99% of the comparisons rendered log LR values > 0 for true contributors, and of these all but one gave log LR > 5, regardless of the number of donors or whether the smallest contributor donated less than 20% of the cells, greatly expanding the collection of cases for which DNA forensics provides informative results.