Forensic features and genetic background exploration of a new 47-autosomal InDel panel in five representative Han populations residing in Northern China.

BACKGROUND: Insertion/deletion (InDel) analysis plays an indispensable role in human identification, population genetics, and biogeographic research. Profiles of individuals in forensic applications worldwide based on a set of autosomal InDel loci (A-InDels) in human genomes have been widely used over the past few years.
METHODS: The new AGCU InDel 50 Kit contains 47 well-chosen A-InDels, ensuring high discriminatory power, and the 2 Y chromosome InDel loci (Y-InDels) are used for sex determination in case of allele dropout at Amelogenin. In this study, five Northern Han populations residing in different geographic areas of China were recruited and genotyped using the assay.
RESULTS: After Bonferroni correction, all 47 A-InDels were in accordance with the lack of significant departures of Hardy-Weinberg equilibrium in all loci and investigated groups. The combined probability of discrimination and the probability of exclusion in the Han population range from 1-3.2240 × 10-19 to 1-1.3030 × 10-19 and 0.9997, respectively. A comprehensive population genetic relationship investigation between Han Chinese and 26 worldwide populations based on allele frequency correlation was carried out. Our results revealed no significant genetic differentiation in Chinese Han groups. Hierarchical clustering, phylogenetic relationship reconstructions, multidimensional scaling, principal component analysis, and structure analysis were performed, and the results indicated that, genetically, Han populations are closely related to East Asians.
CONCLUSION: Overall, this novel 47 A-InDel assay is a valuable tool that could potentially be used for forensic identification and parentage tests.

BACKGROUND

In forensic genetics, autosomal short tandem repeat (STR) genotyping is the main approach in personal identification and paternity testing (Romsos & Vallone, 2015). However, the large size of the PCR amplicons limits analysis of degraded DNA samples (Mills et al., 2006; Weber et al., 2002). Insertion–deletion polymorphisms are a special type of biallelic length polymorphism marker with a low mutation rate, short amplicon size, and wide distribution in the genome (Fondevila et al., 2012; Pereira et al., 2009; Zaumsegel, Rothschild, & Schneider, 2013; Zhao et al., 2017). Since InDels complement the disadvantages of STRs and have gained considerable attention from forensic researchers, various commercialized strategies have been developed in an attempt to supplement forensic investigations. For intelligence purposes, the absence of stutter peaks and the requirement of simple analytical procedures make the amplified products easy to separate using the capillary electrophoresis platform found in most forensic genetic laboratories around the world (Zhao et al., 2018).

Numerous population studies have been conducted with the first multiplex amplification InDel system (Investigator® DIPplex Kit, Qiagen) in recent decades (Jian et al., 2019; Li, Ye, Song, Wang, & Hou, 2019; Zhang et al., 2019; Zhu et al., 2018). In the process of InDel development, a novel multiplex panel named the AGCU InDel 50 Kit can simultaneously amplify as many as 47 advanced A‐InDels, Amelogenin, and 2 Y‐InDels to correctly determine sex (Chen et al., 2019). Notably, the new InDel‐based multiplexes include a large number of overlapping autosomal markers (n = 45) from the 1,000 Genomes Project, more than the DIPplex Kit (n = 17) (Genomes Project et al., 2010, 2012, 2015). Several validation studies have demonstrated its powerful ability and potential in forensic investigations and have shown high informativity among populations (Pan et al., 2019; Wang, Wang, He, Zhu, & Hou, 2019). Han Chinese, accounting for approximately 20% of the total population worldwide, is the most widely distributed ethnic group, with a population size of nearly 1.4 billion (Xu et al., 2009). In our published article, genetic differentiation was observed between Northern and Southern Han Chinese populations based on Y chromosome markers (Lang et al., 2019), but due to the lack of more types of DNA marker‐based data among Northern Han populations from different geographic regions, forensic efficiency has been somewhat restricted. Moreover, the InDel‐based genetic relationship between the Northern Han Chinese population and the adjacent groups or the population among the inter‐Northern Han Chinese populations is still unclear.

In the present study, five Han Chinese populations recruited from five different geographic areas of Northern China (Beijing, Henan, Heilongjiang, Shandong, and Shanxi) were first genotyped by the new generation of the AGCU InDel 50 Kit to fill the geographic sampling gap, aiming to make it more suitable for personal identification and give a comprehensive understanding of the population background in Northern Han Chinese. Regarding the interpopulation comparison, investigation among 2,705 individuals from 26 worldwide populations from the 1000 Genomes Project on the basis of 45 overlapping InDels was performed. The abbreviations and details of the relative populations are listed in Table S1. Our study also provided a preliminary Northern Han database based on these 50 InDel loci.

METHODS

Sample preparation and ethics statement

A total of 535 samples from unrelated Chinese Han individuals were collected from five geographic areas (Figure 1) of China (BJH, Beijing Han Chinese, n = 108; HNH, Henan Han Chinese, n = 110; HLJH, Heilongjiang Han Chinese, n = 114; SH, Shandong Han Chinese, n = 108; SXH, Shanxi Han Chinese, n = 95) with the approval of the Ethics Committee at the Institute of Forensic Medicine, Sichuan University (K2019017). All participants provided informed consent. Genomic DNA was extracted from bloodstains using the standard Chelex‐100 method.

Figure 1

The geographic positions of the five studied Northern Han populations in China

Quality control

DNA 9948 was used as an amplification positive control, and ddH2O was routinely used as a negative control for the whole performance detection of this multiplex assay. Our laboratory has been approved by the China National Accreditation Service for Conformity Assessment (CNAS) and accredited according to quality control measurements (ISO 17025). We strictly followed the requirements and recommendations advocated by the International Society of Forensic Genetics (ISFG) in each batch of experiments.

PCR amplification and genotyping

The PCR amplifications of the InDel markers were simultaneously performed using the AGCU InDel 50 Kit according to the manufacturer's protocol. PCR was conducted on a ProFlex 96‐Well PCR System (Thermo Fisher Scientific, MA, USA) in a total volume of 10 μl, which contained 4 μl of 2.5 × PCR buffer mixture, 2 μl of primer mix, 0.4 μl of heat‐activated Taq polymerase (5 U/μl), 2.6 μl of ddH2O water, and 1.0 μl of template DNA. The amplification setup and cycling parameters consisted of an initial denaturation step at 95°C for 2 min, 29 cycles of 30 s at 94°C, 60 s at 60°C, and 60 s at 72°C, followed by a final extension step for 30 min at 60°C. The amplified products were subsequently detected using capillary electrophoresis on the ABI 3500 Genetic Analyzer (Applied Biosystems), and allele assignment of each locus was performed using GeneMapper ID‐X software (Applied Biosystems).

Statistical analysis

Forensic statistical parameters and allelic frequency distribution were calculated separately for each investigated population in this study. Corresponding population statistical parameters of 47 A‐InDels, such as insertion allele frequencies (DIP+), discrimination power (PD), match probability (MP), probability of exclusion (PE), observed heterozygosity (Ho), and polymorphism information content (PIC), were performed using Powerstats V12 (Promega). Hardy–Weinberg equilibrium and linkage disequilibrium (LD) were calculated using the STRAF online tool (http://www.cmpg.iee.unibe.ch/services/shiny/index_eng.html). The significance level was adjusted by the Bonferroni procedure (p = .05/47 of Hardy–Weinberg equilibrium; p = .05/108 in BJH, p = .05/110 in HNH, p = .05/114 in HLJH, p = .05/108 in SH and p = .05/108 in SXH of LD). The heat map of hierarchical clustering on the basis of + DIP was carried out by the “Package pheatmap” of R principal component analysis (PCA) based on raw data was conducted by Python.

Comprehensive establishment of the genetic relationships between the studied Han populations and reference populations was accomplished through 45 overlapping InDels (rs34421865, rs3076465, rs34287950, rs151335218, rs5787309, rs67405073, rs71852971, rs769299, rs79225518, rs140683187, rs140323077, rs67700747, rs35267904, rs10607699, rs35309403, rs145577149, rs139934789, rs35453727, rs5897566, rs35464887, rs67100350, rs538690481, rs67939200, rs66739142, rs3217112, rs34209360, rs66477007, rs142221201, rs67426579, rs145941537, rs67365630, rs3029189, rs10558392, rs3834231, rs61490765, rs72085595, rs67487831, rs66595817, rs35065898, rs34529638, rs67264216, rs3067397, rs11277697, rs33971783, and rs34419736) from the 1000 Genomes Project. To visualize the genetic similarities and differences among the 31 overall worldwide populations, Arlequin statistical software v3.5 (Excoffier & Lischer, 2010) was used to assess the pairwise genetic distances (Fst). The comparison among different populations for the 45 overlapping InDels was performed via multidimensional scaling plots (MDS) by SPSS Software v21.0 (Saitou & Nei, 1987). Subsequently, a neighbor‐joining tree using MEGA Software (v7.0) was also constructed by means of pairwise Fst values (Kumar, Stecher, & Tamura, 2016). Finally, the assessment of population structures with STRUCTURE version 2.3.4 was run with k ranging from 2 to 6 (Falush, Stephens, & Pritchard, 2007).

RESULTS AND DISCUSSION

Forensic parameters in five Northern Han populations

Allen frequency distributions and other forensic parameters of the 47 A‐InDels in the five investigated Northern Han populations were calculated and are listed in Table S2. The insertion allele frequencies at the 47 A‐InDels ranged from 0.3060 (rs72085595) to 0.7500 (rs67700747) in BJH, 0.2500 (rs3067397) to 0.7410 (rs67700747) in HNH, 0.3200 (rs72085595) to 0.7680 (rs67700747) in HLJH, 0.2500 (rs72085595) to 0.801 (rs67700747) in SH, and 0.2740 (rs67264216) to 0.7890 (rs67700747) in SXH. No significant deviations from Hardy–Weinberg equilibrium were observed after Bonferroni correction. All InDels analyzed were in LD, except rs71852971 and rs67264216 for HNH, even after Bonferroni correction (Table S3). This departure may be due to random statistical effects.

None of the InDel markers with Ho were less than 0.3 in these studied Northern Han groups. The Ho values ranged from 0.3704 (rs10607699 and rs67700747) to 0.5833 (rs79225518) in BJH, 0.3455 (rs10607699) to 0.6182 (rs79225518) in HNH, 0.3070 (rs67700747) to 0.5614 (rs67426579) in HLJH, 0.3241 (rs67700747) to 0.5926 (rs1160980) in SH, and 0.3158 (rs67700747) to 0.5789 (rs60575667) in SXH. All 47 InDels showed PD values greater than 0.5 except rs67700747 in SH (PD = 0.4854) and SXH (PD = 0.4986). Moreover, the PE values for the five Northern Han population groups varied from 0.0969 (rs10607699) to 0.2714 (rs79225518) in BJH, 0.0840 (rs10607699) to 0.3133 (rs79225518) in HNH, 0.0665 (rs67700747) to 0.2471 (rs67426579) in HLJH, 0.0739 (rs67700747) to 0.2821 (rs1160980) in SH, and 0.0702 (rs67700747) to 0.2664 (rs5787309) in SXH. Additionally, the combined power of discrimination (CPD) ranged from 1–3.2240 × 10–19 to 1–1.3030 × 10–19, and the combined power of paternity exclusion (CPE) reached 0.9997. These aforementioned data revealed that all 47 A‐InDels could be a powerful and potential tool for forensic personal identification and paternity testing in Northern Han populations of China.

In brief, concerning the prospects of forensic applications for this new panel, we found it was sufficient for forensic investigations in five representative Han populations residing in Beijing, Henan, Heilongjiang, Shandong, and Shanxi of Northern China. These markers are comparatively useful to serve as supplementary resources in complicated cases.

Hierarchical clustering

A heat map (Wilkinson & Friendly, 2009) of insertion allele frequencies for the 45 overlapping loci was generated by the hierarchical clustering method between the five Northern Han populations analyzed in this study and 26 worldwide populations from the 1000 Genomes Project (Genomes Project et al., 2010, 2012, 2015). As shown in Figure 2, the color of each block ranged from blue to red, representing the lowest insertion allele frequency to the highest insertion allele frequency. All 31 populations were divided into five groups by different colors: African, American, East Asian, European, and South Asian. Hierarchical clustering analysis was generated on the top of the heat map, and five investigated clusters were easily observed. It is suggested that most of them present differences in the insertion allele frequencies between geographically distant population groups. For distributions of insertion allele frequencies, the SH and SXH groups and the BJH and HNH groups were found to share more analogical insertion allele frequency distributions, while HLJH was slightly different than them. Furthermore, HLJH was expected to be distinctive since the population is geographically far from the other four populations in our study, which meant a slightly different genetic structure among HLJH and these Chinese populations.

Figure 2

Heatmap of insertion allele frequencies between 5 studied populations and other 26 populations by Hierarchical clustering method

Population comparisons by Fst values

To further explore the population relationship and genetic similarities among worldwide populations, a comprehensive pairwise Fst analysis was conducted. Pairwise Fst values of a total of 31 populations are listed in Table S4. Then, MDS and phylogenetic relationship reconstruction on the basis of Fst values were performed.

First, genetic distances between five Northern Han populations and other populations were graphically represented by four MDS plots. As shown in Figure 3, four layers of comparisons were completed. In the first layer, we aimed to describe the population relationship and genetic distance at the global level on the basis of the 1000 Genomes Project. Five studied Northern Han populations were clearly clustered in the fourth quadrant, together with East Asian populations in Figure 3a. In the second layer, only 10 East Asian populations were selected for comparison (Figure 3b). Eight Chinese groups were observed to gather relatively close together but were far from non‐Chinese populations (KHV and JPT), which meant a close genetic relationship among these Chinese populations. In the third layer, Figure 3c shows the genetic distance between Chinese Han groups within the same population. Concerning the genetic relationships among different Han populations, the distribution seemed to become more diffuse. Finally, HNH and SH were included in the first quadrant, and HNH and SXH were included in the fourth quadrant. Located in Northeast China, HLJH is farther away than the aforementioned four Han populations and fell into the second quadrant separately in Figure 3d. We revealed that all studied Han populations were concentrated near each other and were relatively distinguished according to geographic distance (HLJH). MDS plots also validated that Han groups were in close genetic relation to East Asian populations, especially Chinese populations.

Figure 3

The MDS showing the genetic relationships between 5 studied Northern Han populations and other reference populations. (a) Respectively, constructed on the basis of 5 Northern Han populations and 26 worldwide populations of 1000 Genomes Project. (b) Respectively, constructed on the basis of 5 Northern Han populations and East Asian populations of 1000 Genomes Project. (c) Respectively, constructed on the basis of 5 Northern Han populations and Chinese populations of 1000 Genomes Project. (d) Respectively, constructed on the basis of 5 Northern Han populations

Afterward, a neighbor‐joining phylogenetic tree based on Fst values was also constructed to evaluate the consistency of MDS. As shown in Figure 4, BJH and HNH and SH and SXH were clustered and clustered with HLJH and then gathered together with other East Asian populations. Five studied Northern Han populations were close to each other and almost located in the Northern Han cluster and separated with Southern groups (CHS, CDS, and JPT). The result of the N‐J tree among 31 populations was in agreement with the above population analysis. That is, all Han populations in this study had a relatively close genetic relationship with East Asian populations residing in the north.

Figure 4

Neighbor‐joining tree between 5 studied Northern Han populations and 26 reference populations of 1000 Genomes Project

PCA and structure using raw data

The two‐dimensional pattern PCA was performed by Python. Figure 5a shows three distinct clusters along PC1 (7.35%): the African circle, one predominantly circle composed of American, European, and South Asian, and the East Asian circle. Moreover, as found previously, Africa was clearly divided from others owing to its special genomes. Additionally, two dominant clusters were observed along PC1 (6.69%): BJH, HLJH, HNH, SH, and SXH combined and clustered on the left side, while CHB and CHS overlapped each other on the right side in Figure 5b. The results indicated that Northern Han populations were close to each other and had a close distance to other East Asian populations. It is visible that PCA based on raw data was in accordance with previous MDS contrast along the horizontal principal component (PC1) and vertical delineation (PC2).

Figure 5

The 2D‐pattern PCA between 5 studied Northern Han populations and reference populations of 1000 Genomes Project by Python. (a) Respectively, constructed on the basis of 5 Northern Han populations and 26 worldwide populations of 1000 Genomes Project. (b) Respectively, constructed on the basis of 5 Northern Han populations and East Asian populations of 1000 Genomes Project

Structure analysis was further performed in our study to mirror the population structure and genetic relationships between five studied Northern Han groups and the other 26 populations with the number of hypothetical populations (K) defined at 2–7. As demonstrated in Figure 6, the optimum K value (Earl & Vonholdt, 2012) was estimated to be K = 3 using STRUCTURE HARVESTER (http://taylor0.biology.ucla.edu/structureHarvester/) by successfully and precisely capturing the population structure. At K = 3, East Asian, African and a third cluster combining European, American, and South Asian populations could be differentiated by distinct discrepancies in color compositions. When K was 4–7, two American populations (MXL and PEL) were distinguished, and no further detailed substructures were dissected for the increase in K values. It followed that the studied Northern Han populations and other populations of East Asia exhibited a similar component cluster pattern, whereas the five studied populations could not be distinguished by their geographical locations. The structural evidence identically revealed that frequent genetic flow might exist between Northern Han populations, which supported the close genetic distance and close geographic connection with each other.

Figure 6

Structure of 5 Northern Han populations with 26 reference worldwide populations varying from 2 to 7. The optimum K value was 3

CONCLUSION

In conclusion, we reported the 47 A‐InDel allelic frequencies and forensic parameters of five Northern Han populations, which exhibited high polymorphism and powerful discrimination capability and can be utilized in forensic identification and parentage tests. Additionally, the population comparison showed that five studied Northern Han populations shared more genetic similarities with East Asians. In summary, the population data in this study could be valuable in enriching genetic databases and provide reference data for forensic tracing and population genetic studies in the future.

CONFLICT OF INTEREST

No potential conflict of interest was reported by the authors.

AUTHOR CONTRIBUTIONS

YPH and FS conceived the idea for the study. ML and HBL performed or supervised laboratory work. LYL analyzed the data. FS and LYL wrote and edited the manuscript. All authors reviewed the final manuscript.

Table S1‐S4

Click here for additional data file.