Development of a SNP-based panel for human identification for Indian populations

The widely employed short tandem repeat (STR)-based panels for forensic human identification (HID) have limitations while dealing with challenging forensic samples involving DNA degradation, resulting in dropping-out of higher molecular weight alleles/loci. To address this issue, bialleic markers like single nucleotide polymorphisms (SNPs) and insertion-deletions (indels), which can be scored even when the template DNA is heavily degraded (<100 bp), have been suggested as alternative markers for HID testing. Recent studies have highlighted their utility in forensic HID and several panels based on biallelic markers have been described for worldwide populations. However, there has been very little information about the behavior of such DNA markers in Indian populations, which is known to possess great genetic diversity. This study describes a two-step approach for designing a SNP-based panel consisting of 70 SNPs for HID testing in Indian populations. In the first step, candidate SNPs were shortlisted from public databases by screening them for several criteria including allelic distribution, genomic location, potential phenotypic expression or functionality and species specificity. The second step involved genotyping the shortlisted SNPs in various Indian populations followed by shortlisting of the best performers for identity-testing. Starting with 592,652 SNPs listed in Human660W-Quad Beadchip (Illumina Inc.), we shortlisted 275 candidate SNPs for identity-testing and genotyped them in 462 unrelated individuals from different population groups in India. Post genotyping and statistical analyses based on biogeographic regions, 206 SNPs demonstrated desired allelic distribution (Heterozygosity ≥ 0.4 and F_ST ≤ 0.02), from which 2–4 widely separated (>20 Mb apart) SNPs from each chromosome were finally selected to construct a panel of 70 SNPs. This panel on average possessed match probability 10e-29 and probability of paternity of 0.99999997, which was orders of magnitude higher than most of the currently employed STR-based chemistries and SNP-based panels that were proposed previously for HID testing. For comparison purpose, genotyping previously reported SNPs for HID in our samples led us to conclude that the panel developed in this study is much more efficient and robust and better suited for the Indian populations.     

A forensic population database in El Salvador: 58 STRs and 94 SNPs

We have genotyped the 58 STRs (27 autosomal, 24 Y-STRs and 7 X-STRs) and 94 autosomal SNPs in Illumina ForenSeq™ Primer Mix A in a sample of 248 men and 143 women from El Salvador, Central America. Regional division (Centro, Oriente, Occidente) showed in almost all cases F_ST values not significantly different from 0, and further analyses were applied only to the undivided, country-wide population. The overall random match probability (RMP) decreased from 6.79 × 10⁻³¹ in length-based genotypes in the 27 autosomal STRs to 1.47 × 10⁻³⁴ in repeat-sequence based genotypes. Combining the autosomal loci in this set, RMP reaches 2.97 × 10⁻⁷⁰. In a population genetic analysis, El Salvador showed the lowest F_ST values with US Hispanics both for autosomal and X-STRs; however, it was much closer to Native Americans for the latter than for the former, in accordance with the well-known gender-biased admixture that created most Latin American populations.     

Analysis of 49 autosomal SNPs in three ethnic groups from Iran: Persians,Lurs and Kurds

A total number of 149 individuals from Iran (Persians, Lurs and Kurds) were analyzed for 49 autosomal SNPs using PCR, SBE and capillary electrophoresis. No deviation from Hardy–Weinberg expectations was observed. One SNP pair (rs1015250–rs251934) showed significant linkage disequilibrium in Kurds. However, this was most likely due to chance. High intrapopulation variability and no significant population structure were observed among the three ethnic groups from Iran. Pairwise F_ST values obtained from the mean numbers of pairwise differences between SNP profiles were calculated for Persians, Lurs, Kurds and eighteen other worldwide populations. For each of the three Iranian ethnic groups, the lowest F_ST values calculated between an Iranian and non-Iranian populations were observed between Iranians and populations in Iraq and Turkey. The three Iranian ethnic groups grouped together with other West Asian populations in the MDS plot drawn from the F_ST values. Statistical parameters of forensic interest calculated for the Iranian ethnic groups showed values of the same order of magnitudes as those obtained for Asians. The mean match probability calculated for the 49 SNPs ranged from 1.7 x 10⁻¹⁸ for Kurds to 1.3 x 10⁻¹⁹ for Persians. Despite the low level of genetic structure observed among Persians, Lurs and Kurds, a single autosomal SNP database should be used with care when extending its forensic application to other Iranian ethnic groups.     

Analysis of 49 autosomal SNPs in an Iraqi population

Forty-nine of the 52 autosomal single nucleotide polymorphisms (SNPs) in the SNPforID 52plex were typed in 101 unrelated Iraqis living in Denmark. No significant deviation from HWE was found in all but one of the 49 SNP systems and no significant pairwise linkage disequilibrium was observed for any SNP pair. When 18 worldwide populations were compared (including populations in Iraq, Turkey, Israel, Pakistan, India, China, Taiwan, Japan, Siberia, Algeria, Somalia, Uganda, Mozambique, Angola, Nigeria, Denmark, Portugal, Spain), a significant global F_ST value was obtained. All but six F_ST values were statistically significant when pairwise comparisons were performed between the 18 populations. The Iraqi population did not show significant difference from the population in Turkey and it grouped together with other Middle-Eastern populations when a multidimensional scaling plot was drawn based on the pairwise F_ST values. The combined mean match probability and the typical paternity index for trios were 8.3 × 10⁻²⁰ and 259,000, respectively, for the Iraqi population.     

Performance of ancestry-informative SNP and microhaplotype markers

The use of microhaplotypes (MHs) for ancestry inference has added to an increasing number of ancestry-informative markers (AIMs) for forensic application that includes autosomal single nucleotide polymorphisms (SNPs) and insertions/deletions (indels). This study compares bi-allelic and tri-allelic SNPs as well as MH markers for their ability to differentiate African, European, South Asian, East Asian, and American population groups from the 1000 Genomes Phase 3 database. A range of well-established metrics were applied to rank each marker according to the population differentiation potential they measured. These comprised: absolute allele frequency differences (δ); Rosenberg’s informativeness for (ancestry) assignment (I_n); the fixation index (F_ST); and the effective number of alleles (A_e). A panel consisting of all three marker types resulted in the lowest mean divergence per population per individual (MDPI = 2.16%) when selected by I_n. However, when marker types were not mixed, MHs were the highest performing markers by most metrics (MDPI < 4%) for differentiation between the five continental populations.     

The effects of Asian population substructure on Y STR forensic analyses

A total of 3046 males of Chinese, Malay, Thai, Japanese, and Indian population affinity were previously typed for the Y STR loci DYS19, DYS385 (counted as two loci), DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, DYS437, DYS438, DYS439, DYS456, DYS458, DYS635, DYS448, and Y GATA H4 using the AmpFlSTR® Yfiler™ kit. These samples were assessed for population genetic parameters that impact forensic statistical calculations. All population samples were highly polymorphic for the 16 Y STR markers with the marker DYS385 being the most polymorphic, because it is comprised of two loci. Most (2677 out of a total of 2806 distinct haplotypes) of the 16 marker haplotypes observed in the sample populations were represented only once in the data set. Haplotype diversities were greater than 99.57% for the Chinese, Malay, Thai, Japanese, and Indian sample populations. For the Y STR markers, population substructure correction was considered when calculating the rarity of a Y STR profile. An F_ST value, rather than a R_ST value, is more appropriate under a forensic model. Because the F_ST values are very small within the Asian populations, the estimate of the rarity of a haplotype comprised of 10–16 markers does not need substructure correction. However haplotypes with fewer markers may require F_ST corrections when calculating the rarity of the profile.     

Population genetic data for 12 X-STR loci in the Northern Han Chinese and StatsX package as tools for population statistics on X-STR

The X-chromosomal short tandem repeat (X-STR) has the advantage in female traces identification against male contamination and in complex kinship cases. In this study, a total of 516 Northern Han Chinese were genotyped using the Investigator^® Argus X-12 QS Kit and pairwise genetic distances (F_ST) were measured between Northern Han Chinese and 35 published populations using the same 12 X-STR loci in four presumed linkage groups (LG). Meanwhile, the StatsX package was compiled to aid the analysis of population statistics on X-STR. Off-ladder alleles were investigated by Sanger sequencing or next generation sequencing (NGS). The results showed that high combined PD_M, PD_F, MEC_Krüger, MEC_Kishida, MEC_Desmarais and MEC_{Desmarais Duo} based on allele frequencies were achieved as 0.999999998, >0.999999999, 0.999996425, 0.999999993, 0.999999993 and 0.999998732, respectively, so did they based on haplotype frequencies. Averaged F_ST and multidimensional scaling (MDS) plot generally mirrored with the biogeography distribution of the studied populations and their historical relationships. A total of 16 unique off-ladder alleles were observed in this experiment, four of which have not been reported yet. The StatsX package could obtain full concordance with established software. Overall, the Investigator^® Argus X-12 QS Kit may provide high polymorphic information for forensic identification and kinship analysis in the Northern Han Chinese population, and the StatsX package can make the workflow smoother for researchers to do population statistical analysis on X-STR.     

FST estimates of 94 populations in China based on STR markers

The proper assessment of DNA evidence in cases of personal identification is a recurring theme in forensics. It is common practice to evaluate the strength of DNA evidence using the likelihood ratio (LR). The accurate use of population allele frequencies is a crucial problem in LR calculation. Allele frequency differences among different populations could be estimated by the F_ST values. Thus, F_ST would also affect LR values by correcting the allele frequencies. In this study, Chinese population allele frequency data were selected from population reports published in Chinese and English journals. The population-specific F_ST values of each population, the overall F_ST values of each province, each region, and the whole country, and the locus-specific F_ST values of different loci were calculated. The LRs using different allele frequencies and different F_ST values were compared based on the combination of simulated genotypes. As a result, the F_ST values of 94 populations, 19 provinces, 7 regions, and the whole country were obtained. The LR was overestimated using allele frequencies of the combined population containing multiple populations rather than using allele frequencies of a population, and the LRs after F_ST correction were lower than those without correction. Conclusively, the correction in conjunction with corresponding F_ST values can make the LRs more accurate and reasonable.     

The genetic structure of native Americans in North America based on the Globalfiler® STRs

Current forensic STR databases, such as CODIS, lack population genetic data on Native American populations. Information from a geographically diverse array of tribes is necessary to provide improved statistical estimates of the strength of associations with DNA evidence. The Globalfiler® STR markers were used to characterize the genetic structure of ten tribal populations from seven geographic regions in North America, including those not presently represented in forensic databases. Samples from the Arctic region, Baja California, California/Great Basin, the Southeast, Mexico, the Midwest, and the Southwest were analyzed for allele frequencies, observed and expected heterozygosities, and F-statistics. The tribal samples exhibited an F_ST or θ value above the conservative 0.03 estimate recommended by the National Research Council (NRC) for calculating random match probabilities among Native Americans. The greater differentiation among tribal populations computed here (θ = 0.04) warrants the inclusion of additional regional Native American samples into STR databases.     

Application of the new insertion–deletion polymorphism kit for forensic identification and parentage testing on the Czech population

Insertion-deletion polymorphisms (INDELs) are diallelic markers derived from a single mutation event. Their low mutation frequency makes them suitable for forensic and parentage testing. The examination of INDELs thus combines advantages of both short tandem repeats (STR) and single nucleotide polymorphisms (SNP). This type of polymorphisms may be examined using as small amplicon size as SNP (about 100 bp) but could be analyzed by techniques used for routine STR analysis. For our population study, we genotyped 55 unrelated Czech individuals. We also genotyped 11 trios to analyze DIPplex Kit (QIAGEN, Germany) suitability for parentage testing. DIPplex Kit contains 30 diallelic autosomal markers. INDELs in DIPplex Kit were tested with linkage disequilibrium test, which showed that they could be treated as independent markers. All 30 loci fulfill Hardy–Weinberg equilibrium. There were several significant differences between Czech and African populations, but no significant ones within European population. Probability of a match in the Czech population was 1 in 6.8?×?10¹²; combined power of discrimination was 99.9999999999%. Average paternity index was 1.13–1.77 for each locus; combined paternity index reached about 27,000 for a set of 30 loci. We can conclude that DIPplex kit is useful as an additional panel of markers in paternity cases when mutations in STR polymorphisms are present. For application on degraded or inhibited samples, further optimization of buffer and primer concentrations is needed.     

Polish population data on 15 autosomal STRs of AmpFlSTR NGM PCR kit

The objective of the research was to provide a comprehensive database of autosomal microsatellite loci included in AmpFlSTR NGM PCR kit for a population of Poland considering possible genetic differentiation of a forensic interest. Fifteen STR markers were analyzed in 2041 unrelated individuals residing in eight geographically different regions. All the loci were found to be in Hardy–Weinberg equilibrium. The combined probability of match is 3.52 × 10⁻¹⁹ and the combined Power of Exclusion is 0.9999998. The F_ST estimate over all 15 STRs is 0.0051 for the Polish population. We established that a combined NGM database may be employed for a Polish population.     

Forensic and population genetic characteristics of 62 X chromosome SNPs revealed by multiplex PCR and MALDI-TOF mass spectrometry genotyping in 4 North Eurasian populations

X chromosome genetic markers are widely used in basic population genetic research as well as in forensic genetics. In this paper we analyze the genetic diversity of 62 X chromosome SNPs in 4 populations using multiplex genotyping based on multi-locus PCR and MALDI-TOF mass spectrometry, and report forensic and population genetic features of the panel of X-linked SNPs (XSNPid). Studied populations represent Siberian (Buryat and Khakas), North Asian (Khanty) and Central Asian (Kazakh) native people. Khanty, Khakas and Kazakh population demonstrate average gene diversity over 0.45. Only East Siberian Buryat population is characterized by lower average heterozygosity (0.436). AMOVA analysis of genetic structure reveals a relatively low but significant level of genetic differentiation in a group of 4 population studied (F_ST = 0.023, p = 0.0000). The XSNPid panel provides a very high discriminating power in each population. The combined probability of discrimination in females (PDf) for XSNPid panel ranged between populations from 0.99999999999999999999999982 in Khakas to 0.9999999999999999999999963 in Buryats. The combined discriminating power in males (PDm) varies from 0.999999999999999792 to 0.9999999999999999819. The developed multiplex set of X chromosome SNPs can be a useful tool for population genetic studies and for forensic identity and kinship testing.     

Genetic population study of Y-chromosome markers in Benin and Ivory Coast ethnic groups

Ninety-six single nucleotide polymorphisms (SNPs) and seventeen short tandem repeat (STRs) were investigated on the Y-chromosome of 288 unrelated healthy individuals from populations in Benin (Bariba, Yoruba, and Fon) and the Ivory Coast (Ahizi and Yacouba). We performed a multidimensional scaling analysis based on F_ST and R_ST genetic distances using a large extensive database of sub-Saharan African populations. There is more genetic homogeneity in Ivory Coast populations compared with populations from Benin. Notably, the Beninese Yoruba are significantly differentiated from neighbouring groups, but also from the Yoruba from Nigeria (F_ST > 0.05; P < 0.01). The Y-chromosome dataset presented here provides new valuable data to understand the complex genetic diversity and human male demographic events in West Africa.     

Thirty autosomal insertion-deletion polymorphisms analyzed using the Investigator® DIPplex Kit in populations from Iraq,Lithuania, Slovenia,and Turkey

Thirty autosomal insertion-deletion (InDel) polymorphisms were analyzed in four populations from Iraq, Lithuania, Slovenia, and Turkey using the commercial kit Investigator® DIPplex. Genotyping issues were encountered for five of the 30 InDels. They were most probably caused by polymorphisms located in the primer binding sites. Population and forensic parameters were calculated. No significant deviations from Hardy-Weinberg equilibrium or significant linkage disequilibrium were detected. The observed heterozygosities ranged from 33% to 61% depending on the marker and the population. The combined probability of exclusion for the 30 markers was 99.7% in all four populations and the matching probabilities were 1 in 3–4 × 10¹² individuals. The multidimensional scaling plot drawn from F_ST distances showed a good concordance between the relative position of the 15 populations included in the plot and their geographic locations.     

Next generation sequencing of SNPs using the HID-Ion AmpliSeq™ Identity Panel on the Ion Torrent PGM™ platform

The HID-Ion AmpliSeq™ Identity Panel (the HID Identity Panel) is designed to detect 124-plex single nucleotide polymorphisms (SNPs) with next generation sequencing (NGS) technology on the Ion Torrent PGM™ platform, including 90 individual identification SNPs (IISNPs) on autosomal chromosomes and 34 lineage informative SNPs (LISNPs) on Y chromosome. In this study, we evaluated performance for the HID Identity Panel to provide a reference for NGS-SNP application, focusing on locus strand balance, locus coverage balance, heterozygote balance, and background signals. Besides, several experiments were carried out to find out improvements and limitations of this panel, including studies of species specificity, repeatability and concordance, sensitivity, mixtures, case-type samples and degraded samples, population genetics and pedigrees following the Scientific Working Group on DNA Analysis Methods (SWGDAM) guidelines. In addition, Southern and Northern Chinese Han were investigated to assess applicability of this panel. Results showed this panel led to cross-reactivity with primates to some extent but rarely with non-primate animals. Repeatable and concordant genotypes could be obtained in triplicate with one exception at rs7520386. Full profiles could be obtained from 100 pg input DNA, but the optimal input DNA would be 1 ng–200 pg with 21 initial PCR cycles. A sample with ≥20% minor contributor could be considered as a mixture by the number of homozygotes, and full profiles belonging to minor contributors could be detected between 9:1 and 1:9 mixtures with known reference profiles. Also, this assay could be used for case-type samples and degraded samples. For autosomal SNPs (A-SNPs), F_ST across all 90 loci was not significantly different between Southern and Northern Chinese Han or between male and female samples. All A-SNP loci were independent in Chinese Han population. Except for 18 loci with H_e <0.4, most of the A-SNPs in the HID Identity Panel presented high polymorphisms. Forensic parameters were calculated as >99.999% for combined discrimination power (CDP), 0.999999724 for combined power of exclusion (CPE), 1.390 × 10¹¹ for combined likelihood ratio (CLR) of trios, and 2.361 × 10⁶ for CLR of motherless duos. For Y-SNPs, a total of 8 haplotypes were observed with the value of 0.684 for haplotype diversity. As a whole, the HID Identity Panel is a well-performed, robust, reliable and high informative NGS-SNP assay and it can fully meet requirements for individual identification and paternity testing in forensic science.     

Results for five sets of forensic genetic markers studied in a Greek population sample

A population sample of 223 Greek individuals was typed for five sets of forensic genetic markers with the kits NGM SElect™, SNPforID 49plex, DIPplex^®, Argus X-12 and PowerPlex^® Y23. No significant deviation from Hardy–Weinberg expectations was observed for any of the studied markers after Holm–Šidák correction. Statistically significant (P < 0.05) levels of linkage disequilibrium were observed between markers within two of the studied X-chromosome linkage groups. AMOVA analyses of the five sets of markers did not show population structure when the individuals were grouped according to their geographic origin. The Greek population grouped closely to the other European populations measured by F_ST^* distances. The match probability ranged from a value of 1 in 2 × 10⁷ males by using haplotype frequencies of four X-chromosome haplogroups in males to 1 in 1.73 × 10²¹ individuals for 16 autosomal STRs.     

The maternal inheritance of Alto Paraná revealed by full mitogenome sequences

Most studies on maternal lineages of South America populations are restricted to control region (CR) markers and, for some geographical regions, the number of studied samples does not adequately represent the existing diversity. This is the case of mitochondrial DNA (mtDNA) studies on Paraguay that are limited to two Native ethnic groups. To overcome this deficiency, we analysed the mitogenomes from 105 individuals living in Alto Paraná, the second most populated department of the country. Using the Precision ID mtDNA Whole Genome Panel, the molecule was sequenced on Ion S5. The majority of the haplotypes belong to the Native American lineages A, B, C and D. Analyses of maximum parsimony using mitogenome data retrieved from publications and in The 1000 Genomes Project showed a high number of new native American subclades in Paraguay. Also, none of the haplotypes found in Alto Paraná match the remaining South American samples, which include admixed populations from Colombia, Peru and Ecuador, and natives from Colombia and Ecuador. F_ST genetic distance analysis showed that the native genetic background of Alto Paraná has an intermediate position between the Amazonian groups and the admixed populations from Peru and Ecuador, supporting the theory about the Amazonian origin of the Tupi-Guarani and, at the same time, showing the influence of other linguistic groups.     

Population and forensic data for three sets of forensic genetic markers in four ethnic groups from Iran: Persians,Lurs, Kurds and Azeris

A total of 255 individuals (Persians, Lurs, Kurds and Azeris) from Iran were typed for three sets of forensic genetic markers with the NGM SElect™, DIPplex^® and Argus X-12 kits. Statistically significant deviations (P ≤ 0.002) from Hardy–Weinberg expectations were observed for the insertion-deletion markers HLD97 and HLD93 after Holm–Šidák correction. Statistically significant (P < 0.05) levels of linkage disequilibrium were observed between markers within two of the four studied X-chromosomal linkage groups. AMOVA analyses of the three sets of markers did not show population structure when the individuals were grouped according to their ethnic group. The Iranian population grouped closely to populations living geographically near to Iran based on pairwise F_ST distances. The matching probabilities ranged from 1 in 3.2 × 10⁷ males by using haplotype frequencies of four X-chromosomal haplogroups to 1 in 3.4 × 10²¹ individuals for the 16 autosomal STRs.