We collected peripheral blood samples or bloodstains from 262 Tujia healthy and unrelated individuals in Hubei province, China, with prior written informed consents, and then stored them at deep cryopreserve or at room temperature in the dry and ventilated place. Genomic DNA of whole blood samples was further extracted using the Magbead Blood Spots DNA kit (CWBIO, Beijing, China) and quantified using the NanoDrop 2000 instrument (Thermo Fisher Scientific, Waltham, MA, United States) following the recommendations of the manufacturers, and then diluted to about 1ng/ul. A total of 26 reference populations from the 1000 Genomes Project phase 3 database were selected as reference populations for population genetic analyses in this study (population information was shown in Supplementary Table S1). These populations were divided into five continental populations, namely, African (AFR), American (AMR), East Asian (EAS), European (EUR), and South Asian (SAS). The research was approved by the Ethics Committee of Xi’an Jiaotong University (No. 2019-1039).

The AGCU InDel 60 kit is a novel panel with a six-dye multiplex amplification system, which includes 57 autosomal InDels, two Y-chromosomal InDels, and one amelogenin locus. Information (names, chromosomal localizations, and fluorescence labeling) of 60 loci is shown in Figure 1.

Multiplex PCR for 60 InDel loci was performed on the GeneAmp PCR System 9700 Thermal Cycler (Thermo Fisher Scientific, South San Francisco, CA, US) according to the instructions on the AGCU InDel 60 kit. The multiplex PCR system was 25 μL in total which contained 10.0 μL reaction mix, 5.0 μL InDel 60 primers (AGCU ScienTech Incorporation, Wuxi, Jiangsu, China), 1.0 μL U-Taq enzyme, 1.0 ng DNA extracted from whole blood samples or 1mm² bloodstain samples without extraction, and sterilized deionized water. The PCR parameters were as follows: initial denaturation at 95°C for 5 min; then, 28 cycles of denaturation at 94°C for 30s, renaturation at 60°C for 1 min, extension at 62°C for 1 min; and final extension at 72°C for 10 min; and stored at 4°C in the end. After centrifugation at 3000 rpm for 5 min, 1.0 μL amplification products or 1.0 μL InDel 60 Allelic Ladder was mixed with 12.0 μL deionized formamide and 0.5 μL AGCU Marker SIZ-500. The loading mixture was denatured at 95°C for 3 min and then chilled on ice for 3 min immediately. Finally, capillary electrophoresis detection was performed using an ABI 3500xL Genetic Analyzer (Thermo Fisher Scientific, Waltham, Massachusetts, United States), and genotyping was performed using GeneMapper^® ID-X Software version 1.5 (Thermo Fisher Scientific, Waltham, Massachusetts, United States). DNA 9948 and deionized water were used as positive and negative controls for each detection.

The information of 60 InDels was visualized using the R software ggalluvial package. The Hardy–Weinberg equilibrium (HWE) tests of the 57 InDel loci and linkage disequilibrium (LD) analyses of pairwise InDels in the Hubei Tujia group were assessed using STRAF v1.0.5 software (Gou and Alexandre, 2017); in addition, allele frequencies, heterozygosity (Ho), expected heterozygosity (He), polymorphism information content (PIC), discrimination power (PD), probability of paternity exclusion (PE), and probability of match (PM) values were also calculated with this software. Then, the violin map of forensic parameters and the heat map of insertion allele distribution frequencies were drawn using the R software ggplot2 package and pheatmap package, respectively. DISPAN program (Ota, 1993) and Arlequin v3.5 software (Excoffier and Lischer, 2010) were used to obtain the values of the pairwise Nei’s genetic distances (D_A distances) and fixation index (F _ST ) among 27 populations. And phylogenetic tree was constructed based on the pairwise D _A distances using the ggtree package (Yu et al., 2018). Principal component analysis (PCA) was performed using R software based on population insertion allele frequencies among 27 populations. ADMIXTURE software was used to conduct STRUCTURE cluster analysis for all populations, and the pophelper online tool (http://pophelper.com/) was used to draw the visual bar charts.

The p-values for HWE tests at the 57 InDels in the studied Tujia group are shown in Table 1. After the Bonferroni correction, there were no significant deviations from HWE at 57 InDels (p > 0.05/57 = 0.00087719). LD tests among these 57 InDels were also evaluated in the studied Tujia group; all pairwise r ² values were higher than 0.0000313282, indicating that there were no significant LD in these pairs of 57 InDels.

As shown in Table 1, the insertion allele frequencies of the 57 InDels in the studied Tujia group ranged from 0.2901 (rs72085595) to 0.7328 (rs66477007). In order to evaluate the forensic application efficiency of the system of 57 InDels in the studied Tujia group, we calculated the forensic parameters (Ho, He, PIC, PD, PE, and PM). The lowest Ho and He values shown at the locus rs66477007 were 0.3435 and 0.3923, respectively. The rs561160795 locus showed the highest Ho value (0.5573); the rs145010051 and rs67365630 loci showed the highest He value (0.5009). The PIC values of the 57 InDels in the studied Tujia group ranged from 0.3149 (rs66477007) to 0.3750 (rs145010051 and rs67365630). Moreover, the PD, PE, and PM values of the 57 InDels varied from 0.5581 (rs66477007) to 0.6461 (rs145010051), 0.0831(rs66477007) to 0.2427 (rs561160795), and 0.3539 (rs145010051) to 0.4419 (rs66477007) in the studied Tujia group, respectively. The values of combined PD (CPD) and combined PE (CPE) were calculated among the joint application of the 57 InDels, which reached to 0.99999999999999999999999699822 and 0.999975177214539, respectively.

The heatmap based on the insertion allele frequencies of 57 InDels for the studied Tujia group and 26 reference populations is shown in Figure 2. The frequency values were expressed by different color of the cube in the heat map; the color changed from blue to orange as the insertion allele frequency values decreased. As shown in Figure 2, the allele frequency distributions of the studied Tujia group were similar to those of the five reference populations in EAS and showed significant differences from those of other four reference intercontinental populations.

Cluster analysis showed that these 27 populations were mainly divided into two clusters, and seven AFR populations constituted one of the main clusters. As shown in Figure 2, twelve InDels (loci rs145010051, rs113011930, rs76158822, rs60867863, rs77635204, rs46875868, rs79225518, rs57981446, rs77206391, rs34076006, rs66739142, and rs145191158) were high insertion allele frequency distributions (>0.9) in AFR populations. The other main cluster was composed of four continental populations: EUR, SAS, EAS, and AMR. The studied Tujia group in this research was in the same sub-cluster with EAS populations due to their relatively similar insertion allele frequency distributions. Except for loci rs72085595, rs3067397, rs71852971, rs67264216, and rs11277697, the insertion allele frequencies of other loci ranged from 0.4 to 0.7 in EAS populations. In addition, we also found that some loci showed particular allele frequency distributions in different continental populations. For example, the insertion allele frequency of the locus rs72085595 was lower in the studied Tujia group and EAS populations but higher in the AFR and EUR populations. Meanwhile, loci rs145941537 and rs34529639 were higher frequency values in the studied Tujia group and EAS populations but lower frequency values in the AFR and EUR populations.

Based on these 57 InDels, we compared the relevant forensic parameters (He, Ho, PD, PE, PIC, and PM) of the studied Tujia group and five continental populations. As shown in Figure 3, the PIC values of 57 InDel loci in the studied Tujia group and five continental populations ranged from 0.001511 to 0.374979 (AFR), 0.005731 to 0.375 (AMR), 0.307745 to 0.374965 (EAS), 0.005929 to 0.374937 (EUR), and 0.043011 to 0.374983 (SAS). The He values ranged from 0.001513 to 0.500337 (AFR), 0.005755 to 0.500722 (AMR), 0.380288 to 0.500426 (EAS), 0.005952 to 0.500371 (EUR), and 0.044023 to 0.500478 (SAS). The Ho values varied from 0.001513 to 0.537065 (AFR), 0.005764 to 0.544669 (AMR), 0.367063 to 0.559524 (EAS), 0.005964 to 0.518887 (EUR), and 0.0409 to 0.521472 (SAS). The PD values varied from 0.003021 to 0.636413 (AFR), 0.011461 to 0.65417 (AMR), 0.544871 to 0.638314 (EAS), 0.011857 to 0.641076 (EUR), and 0.082368 to 0.640437 (SAS). The PM values ranged from 0.363587 to 0.996979 (AFR), 0.34583 to 0.988539 (AMR), 0.361686 to 0.455129 (EAS), 0.358924 to 0.988143 (EUR), and 0.359563 to 0.917632 (SAS). The PE values ranged from 0.000002282 to 0.222042 (AFR), 0.0000328 to 0.229663 (AMR), 0.09511 to 0.245095 (EAS), 0.00003515 to 0.20456751 (EUR), and 0.001547 to 0.20699 (SAS).

We evaluated the population genetic relationships of the studied Tujia group and 26 reference populations by PCA, structure, D _A , and F _ST analyses. As shown in Figure 4, we utilized two PCA plots to illustrate the genetic relationships among these 27 populations based on 57 InDels at the population level, where the same color represented the same continental populations, however, the studied Tujia group was colored in separate color. The first, second, and third components (PC1, PC2, and PC3) could explain 46.6%, 24.9%, and 8.1% of the total variance, respectively. The AFR and EAS populations could be distinguished from other continental populations on PC1, the EUR populations on PC2, and the SAS populations on PC3. The studied Tujia group was located at the cluster of EAS populations.

STRUCTURE analyses of the studied Tujia group and 26 reference populations at the individual level are shown in Figure 5. At K = 2, the ancestral components of AFR populations were mainly light blue, while those of EAS populations were mainly dark blue. At K = 3, the ancestral components of AFR populations were mainly light blue, those of EAS populations were mainly green, and those of EUR populations were mainly dark blue. With the increase in the K values, the AMR and SAS populations also gradually showed different ancestral compositions compared with AFR, EUR and EAS populations. Regardless of the values of K, the studied Tujia group always showed similar ancestral compositions with the EAS populations.

The phylogenetic tree on the basis of pairwise D _A genetic distances, and F _ST heatmap could visualize the genetic relationships of the studied Tujia group and 26 reference populations. As shown in Figure 6, these 27 populations were divided into two main branches: one was composed of seven AFR populations, while the other consisted of EAS (including the studied Tujia group), SAS, AMR, and EUR populations. Specifically, the studied Tujia group and CHS were clustered firstly, and then incorporated into the CHB and other three EAS populations in turn; all six EAS populations gathered into one branch eventually. The phylogenetic tree and heat map revealed that the genetic relationships between the studied Tujia group and EAS populations were closer than those between the studied Tujia group and AFR populations, correspondingly.