Forensic characteristics and population genetics of Chinese Kazakh ethnic minority with an efficient STR panel

The collection of population samples

Whole blood samples were withdrawn from 501 unrelated healthy individuals of Chinese Xinjiang Kazakh group. During the process of sample collection, we ensured the genetic relationships by inquiring and no blood relationship was existed between any two individuals at least in three generations. Moreover, all individuals should meet the requirements of being the aboriginal inhabitants. The research was officially authorized by the ethics committee of Xi’an Jiaotong University Health Science Center, China (Ethical Application No: 2018-518) before the study, and was performed as per the human and ethical research principles of Xi’an Jiaotong University Health Science Center. All the participators signed informed consent statements prior to the specimen collections.

Multiplex amplification and STR genotyping

DNA was directly amplified without extraction based on a multiplex PCR system that simultaneously amplified 22 STR loci from Microreader™ 23sp ID kit. PCR amplification was implemented on the GeneAmp^® PCR 9700 thermocycler (Applied Biosystems, Foster City, CA, USA) with 25 µL reaction volume, and amplified products were subsequently separated and analyzed on the ABI PRISM^® 3130XL Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) with reference to internal lane standard Org500 (including different length fragments: 50, 75, 100, 139, 150, 160, 200, 300, 340, 350, 400, 450, 490 and 500 bp). The specific reaction system and conditions for electrophoresis were referred to the previously published research (Li et al., 2017). The analyzation of electrophoresis results were revealed by GeneMapper^® ID-X 1.3 software (Applied Biosystems, Foster City, CA, USA).

Population genetic and forensic statistical analyses

For the studied Kazakh group, the distributions of allelic frequencies and forensic statistical parameters containing the observed heterozygosity (Ho), power of exclusion (PE), power of discrimination (PD), polymorphism information content (PIC) and Hardy–Weinberg equilibrium (HWE) of 22 STR loci were calculated by the modified Powerstats software v.1.2 (Tereba, 1999). Expected heterozygosity (He) was calculated as described by Nei (1978). The estimates of linkage disequilibrium (LD) (Slatkin, 2008) of 22 STR loci were performed by Genepop v.4.0.10 (http://genepop.curtin.edu.au/). Correlation coefficients (r²) for all allele combinations were ascertained by SHEsis online tool (Yongyong & Lin, 2005). The locus-by-locus fixation index (Fst values) and corresponding probability (p) values of population genetic differentiations were estimated by Arlequin software v.3.5 (Excoffier & Lischer, 2010) using the analysis of molecular variance (AMOVA) method. The DISPAN program (Ota, 1993) was implemented to calculate Nei’s genetic distance (DA) values (Nei, 1972) utilizing the raw STR genotyping data of Kazakh group. The population pairwise genetic differentiation Fst values and p values were carried out by Arlequin software v.3.5 (Excoffier & Lischer, 2010) based on raw STR genotyping data. The heat maps of population pairwise D_A and Fst values were conducted using R statistical software v3.0.2 (Dean & Nielsen, 2007). The description of population genetic structure was presented by the STRUCTURE software v.2.3.4 (Pritchard, Stephens & Donnelly, 2000) and the optimum K value, the number of hypothetical ancestral populations, was estimated by Structure Harvester v.0.6.94 (Pritchard, Stephens & Donnelly, 2000).

Hardy–Weinberg equilibrium and linkage disequilibrium analyses

HWE tests were performed and the results showed no significant deviations from HWE at these 22 STR loci in Kazakh group after Bonferroni correction (Curtin & Schulz, 1998) (p = 0.05∕22 = 0.00227273). In Fig. S1, the locations of STR loci were indicated on each autosomal chromosome. Furthermore, as shown in Table S1, significant p-values for LD analyses were detected in D10S1435-D15S659 and D13S325-D17S1290 after applying Bonferroni correction (Curtin & Schulz, 1998) (p = 0.05∕231 = 0.00021645). The r² values of 22 STR loci were displayed in the Fig. S2. The result transpired that the r² values of STR pairs were all less than 0.004.

Allelic frequencies and forensic parameters of Kazakh group

Raw genotyping data were presented in the Table S2. Allelic frequencies of the 22 STRs in the Kazakh group were enumerated in Table 1. Totally 276 alleles for 22 loci were found with the corresponding allelic frequencies varied from 0.001 to 0.3623. The least number of alleles (seven) was observed at D4S2366 locus, while the maximum (20) was at D12S391 locus. As presented in Fig. S3, the values of PE and Ho were in the range of 0.4903 to 0.7877, and 0.7385 to 0.8962, respectively, with the minimum values detected at D17S1290 locus and maximum at D20S470 locus. Additionally, the minimum values of PD, He and PIC emerged at D10S1435 locus with 0.8993, 0.7523 and 0.7124, separately. Conversely, the maximum values were all observed at D7S3048 locus with 0.9732, 0.8827 and 0.8700, respectively. The cumulative power of discrimination value for all 22 STR loci in the Kazakh group was 1-1.00E⁻²⁸.

Interpopulation differentiation analyses between Kazakh and reference groups

In order to explore the population genetic differentiations between Kazakh group and the five previously reported reference groups, we calculated the locus-by-locus Fst values with p values of the 22 loci. As shown in Table 2, significant differences were observed between Kazakh and Xinjiang Uygur (Song et al., 2017) at nine loci, Xinjiang Hui (Fang et al., 2018) at 14 loci, Guangdong Han (Chen et al., 2016) at nine loci, Northern Han (Xie et al., 2015) (Hebei, Henan, Shaanxi) at 18 loci and Hainan Li (Chen et al., 2016) at 17 loci. However, after Bonferroni correction (p = 0.05∕110 = 0.00045455), Xinjiang Kazakh had significant differences with Xinjiang Uygur at one loci, Xinjiang Hui at three loci, Northern Han and Hainan Li at ten loci.

The D_A and Fst (p) values of pairwise populations were listed in Tables S3 and S4, respectively. Kazakh group had the shortest distance with Uygur group (D_A = 0.0076), followed by Xinjiang Hui (D_A = 0.0112), Northern Han (D_A = 0.0136), Guangdong Han (D_A = 0.0244) and Hainan Li (D_A = 0.0337) groups. The pairwise Fst values were also similar to the above D_A results: the studied Kazakh group had the lowest Fst value with Xinjiang Uygur group and the highest with Hainan Li group. To visualize the results of population pairwise D_A and Fst values more straightforwardly, we performed two heat maps of D_A and Fst values between Kazakh and other reference groups as presented in Figs. 1A and 1B. The population structure analyses of the Kazakh and five reference groups were performed using the STRUCTURE software with the result depicted in Fig. 2. K = 4 was the most suitable configuration relying on the output posterior probability results (Evanno, Regnaut & Goudet, 2010). Individuals were represented by a vertical line and the colors stood for estimated membership fractions (Rosenberg et al., 2002).