After obtaining written informed consents, blood samples of 221 unrelated healthy individuals were gathered from the Hani group residing in the Yunnan province of China. According to self-reported statements, all the participants included in present study had no common ancestors within three generations, and no migration events occurred in their family history. The sample collections strictly observed the principles of human and ethical research and were approved by the ethics committee of Xi’an Jiaotong University Health Science Center (Approval No. 2019-1039). In the 4008 individuals from 30 reference populations, 26 worldwide populations were from the 1000 Genomes Project Phase 3 database, and the remaining four populations were from previously published studies [14, 15, 16, 17, 19]. The detailed information and corresponding abbreviations for 31 populations are listed in Supplementary Table 1.

The Chelex-100 method was used to extract genomic DNA. The multiplex PCR amplification of the novel panel was conducted on the GeneAmp PCR system 9700 Thermal Cycler (Thermo Fisher Scientific, Foster City, CA, USA) with the corresponding reaction system and amplification condition as previously reported [15]. Subsequent PCR amplification products were separated and detected using the ABI 3500xL Genetic Analyzer (Thermo Fisher Scientific, Foster City, CA, USA). GeneMapper ID-X software v1.5 (Thermo Fisher Scientific, Foster City, CA, USA) was used to implement the InDel genotyping. The positive and negative controls during the experimental procedures were DNA 9947A as well as 9948, and deionized water, respectively.

The allele frequencies and forensic parameter values of autosomal 57 InDels containing the polymorphism information content (PIC), match probability (MP), power of discrimination (PD), probability of exclusion (PE), observed heterozygosity (Hobs), and expected heterozygosity (Hexp) in Chinese Yunnan Hani group and reference East Asian populations were measured by the STRAF online program (version v1.0.5, Gouy & Zieger, Bern, Switzerland) and visualized by raincloud plots by R software (version 4.1.2, R Foundation for Statistical Computing, Vienna, Austria) [20]. The values of cumulative power of discrimination (CPD) and probability of exclusion (CPE) of the 57 autosomal InDels in Chinese Yunnan Hani group were performed by the corresponding formula. The Hardy-Weinberg equilibrium (HWE) tests for the 57 autosomal InDels and linkage disequilibrium (LD) analyses for pairwise InDels in Chinese Yunnan Hani group were also conducted by the STRAF online program (version v1.0.5, Gouy & Zieger, Switzerland) [20]. The pairwise relationship tests involving full siblings and half siblings were simulated by Familias software (version 3, Kling & Tillmar & Egeland, Norwegian, Sweden) based on the allele frequencies of 57 autosomal InDels in Chinese Yunnan Hani group, and the results of likelihood ratio (LR) distributions were plotted with the density plot function via R software [21]. Locus by locus analysis of molecular variance (AMOVA) between Chinese Yunnan Hani group and 30 reference global populations together with pairwise fixation index (F${}_{\text{𝑆𝑇}}$) values among the 31 worldwide populations were analyzed by the Arlequin software (version 3.5.1.2, Excoffier & Lischer, Lausanne, Switzerland) [22]. The D${}_{\text{𝐴}}$ distances of pairwise populations were estimated by the DISPAN program (Nei & Tajima & Tateno, Houston, TX, USA; Saitama, Japan) [23]. Two phylogenetic trees were described serially by the MEGA software (version 7, Tamura, Japan) in view of the D${}_{\text{𝐴}}$ distances with the neighbor-joining method, and the PHYLIP software (version 3.69, Shimada, Japan) with the UPGMA method in light of allele frequencies of 57 autosomal InDels in 31 global populations [24, 25]. The heatmaps of insertion allele frequencies for the 57 autosomal InDels, F${}_{\text{𝑆𝑇}}$ values and D${}_{\text{𝐴}}$ distances among Chinese Yunnan Hani group and 30 reference populations were established through R software. The pairwise F${}_{\text{𝑆𝑇}}$ values of 57 autosomal InDels among pairwise intercontinental populations were calculated using Genepop software (version 4.7, Rousset, France) and presented by a boxplot through R software [26]. Principal component analysis (PCA) plots, including four individual level plots derived from the raw genotyping data and one population level map based on allele frequencies, the cos2 values of 57 autosomal InDels, along with multidimensional scaling (MDS) analysis based on pairwise F${}_{\text{𝑆𝑇}}$ values among Chinese Yunnan Hani group and 30 reference populations were also created by R software. Population genetic structure analysis was carried out by STRUCTURE software (version 2.3.4, Evanno & Regnaut & Goudet, Lausanne, Switzerland) with the hypothetical ancestry clusters (K) values set at 2 to 7 and the number of iterations set at 15 [27]. The optimum K value was estimated through the online tool Structure Harvester (Earl, Santa Cruz, CA, USA) [28]. The average Q-matrices for the 15 iterations were assessed by CLUMPP software (version 1.1.2, Jakobsson, Ann Arbor, MI, USA) [29]. When at the optimum K value, the estimated ancestry components of 31 populations and 4229 individuals were visualized in two circle plots by the AncestryPainter software (Feng & Lu & Xu, Shanghai, China) [30]. The locations of Chinese Yunnan Hani group and 30 reference populations on the world map, the pie plots and boxplots of estimated ancestral components in Chinese Yunnan Hani group and 30 reference populations by assuming three ancestral populations, and the triangle plot represented the estimated ancestry components of 4229 individuals from different intercontinental origins were displayed by R software.

The results of HWE tests and LD analyses of the 57 autosomal InDels in Chinese Yunnan Hani group are presented in Supplementary Tables 2,3, separately. Following the use of the Bonferroni correction, no significant deviation from HWE (p = 0.05/57 = 0.0008771) was observed, and all pairwise InDel loci were in linkage equilibrium status (p = 0.05/1596 = 0.00003133).

The allele frequencies and corresponding forensic parameters of the 57 autosomal InDels in Chinese Yunnan Hani group were recorded in Supplementary Table 2 and depicted by the raincloud chart in Fig. 1A, respectively. Results showed that the minimum insertion allele frequency was observed at the rs72085595 locus (0.2227), while the maximum insertion allele frequency was observed at the rs66595817 locus (0.7820). The values of PIC, MP, PD, PE, Hobs, Hexp in Chinese Yunnan Hani group ranged from 0.2828 (rs66595817) to 0.3750 (rs145191158 and rs145941537), 0.3510 (rs34421865) to 0.4926 (rs66595817), 0.5074 (rs66595817) to 0.6490 (rs34421865), 0.0689 (rs72085595) to 0.2448 (rs67487831), 0.3128 (rs72085595) to 0.5592 (rs67487831), and 0.3418 (rs66595817) to 0.5012 (rs145191158 and rs145941537), respectively. The values of CPD and CPE in Chinese Yunnan Hani group were 0.9999999999999999999999968 and 0.999958, respectively.

Fig. 1.

The raincloud charts of forensic parameters of 57 autosomal InDels. (A) The raincloud chart of forensic parameters of 57 autosomal InDels in Chinese Yunnan Hani group. (B) The raincloud chart of forensic parameters of 57 autosomal InDels in East Asian populations including Chinese Yunnan Hani group.

The allele frequencies and forensic parameters of the 57 autosomal InDels in East Asian populations including Chinese Yunnan Hani group were also calculated to further assess the forensic efficiency of the novel amplification system in East Asian populations. The results are listed in Supplementary Table 4 and exhibited by another raincloud chart in Fig. 1B, respectively. The minimum insertion allele frequency was at the rs72085595 locus (0.2541), and the maximum insertion allele frequency was at the rs10607699 locus (0.7189). The values of PIC, MP, PD, PE, Hobs, Hexp in Chinese Yunnan Hani group and reference East Asian populations ranged from 0.3073 (rs72085595) to 0.3750 (rs35453727), 0.3606 (rs10590825) to 0.4557 (rs72085595), 0.5443 (rs72085595) to 0.6394 (rs10590825), 0.0939 (rs72085595) to 0.2035 (rs146875868), 0.3647 (rs72085595) to 0.5177 (rs146875868), and 0.3792 (rs72085595) to 0.5001 (rs35453727), respectively.

The pairwise relationship tests involving full siblings and half siblings were simulated based on the allele frequencies of 57 autosomal InDels, and the results are shown in Fig. 2A,B, to further evaluate the statistical potency of the 57 autosomal InDels in analyzing complex cases for kinship testing in Chinese Yunnan Hani group. In the results of simulated full sibling tests in Fig. 2A, a small overlap in the distributions of LR values could be observed between full siblings and unrelated individuals. In the results of simulated half sibling tests in Fig. 2B, the observed overlap was further increased in the distributions of LR values between half siblings and unrelated individuals.

Fig. 2.

The results of simulated full sibling and half sibling tests. (A) Log${}_{10}$LR distribution plot to distinguish full siblings from unrelated individuals in Chinese Yunnan Hani group when using 57 autosomal InDels. (B) Log${}_{10}$LR distribution plot to distinguish half siblings from unrelated individuals in Chinese Yunnan Hani group when using 57 autosomal InDels.

The insertion allele frequencies in view of 57 autosomal InDels in the 31 populations are listed in Supplementary Table 5 and visually demonstrated by a heatmap in Fig. 3. The 57 autosomal InDels showed different insertion allele frequency distributions among the five reference intercontinental populations. For example, the rs3076465, rs71852971, rs3067397, rs72031009, and rs34419736 loci exhibited relatively large insertion allele frequencies in African populations but small insertion allele frequencies in other four intercontinental populations. The rs79225518, rs113011930, rs145010051, rs146875868, and rs77635204 loci embodied relatively high insertion allele frequencies in East Asian populations but low insertion allele frequencies in other four intercontinental populations.

Fig. 3.

The heatmap of insertion allele frequencies of 57 autosomal InDels in Chinese Yunnan Hani group and 30 reference populations. ACB, African Caribbean in Barbados; ASW, African Ancestry in Southwest US; ESN, Esan in Nigeria; GWD, Gambian in Western Division, The Gambia; LWK, Luhya in Webuye, Kenya; MSL, Mende in Sierra Leone; YRI, Yoruba in Ibadan, Nigeria; CLM, Colombian in Medellin, Colombia; MXL, Mexican Ancestry in Los Angeles, California; PEL, Peruvian in Lima, Peru; PUR, Puerto Rican in Puerto Rico; CEU, Utah residents with Northern and Western European ancestry; FIN, Finnish in Finland; GBR, British in England and Scotland; IBS, Iberian populations in Spain; TSI, Toscani in Italy; BEB, Bengali in Bangladesh; GIH, Gujarati Indian in Houston, TX; ITU, Indian Telugu in the UK; PJL, Punjabi in Lahore, Pakistan; STU, Sri Lankan Tamil in the UK; CDX, Chinese Dai in Xishuangbanna, China; CHB, Han Chinese in Bejing, China; CHS, Southern Han Chinese, China; JPT, Japanese in Tokyo, Japan; KHV, Kinh in Ho Chi Minh City, Vietnam; SHP, Dingjie Sherpa, China; HNL, Hainan Li, China; HNH, Hunan Han, China; YNM, Yunnan Miao, China; YNH, Yunnan Hani, China.

Following Bonferroni correction, the significance threshold was changed to 0.0012 (p = 0.05/57 = 0.0008771). The locus-by-locus p values computed using the AMOVA approach are listed in Supplementary Table 6. The Yunnan Hani group showed the minimum amount of significant differences at 2–30 loci with East Asian populations, but the maximum number of significant differences at 32–41 loci with African populations. In addition, the significant differences among Chinese Yunnan Hani group and the remaining populations were observed to be 16–30 loci in four American populations, 36–41 loci with five European populations, and 25–27 loci with five South Asian populations.

The pairwise F${}_{\text{𝑆𝑇}}$ values and D${}_{\text{𝐴}}$ distances among Chinese Yunnan Hani group and 30 reference populations are displayed in Supplementary Tables 7,8, respectively. Compared with other populations, Chinese Yunnan Hani group showed the minimum F${}_{\text{𝑆𝑇}}$ value with KHV (0.0090), followed by CHS (0.0095) and CDX (0.0103). The maximum F${}_{\text{𝑆𝑇}}$ value was observed between Chinese Yunnan Hani group and YRI (0.2267), followed by the GWD (0.2236) and ESN (0.2206). The D${}_{\text{𝐴}}$ distances among Chinese Yunnan Hani group and 30 reference populations exhibited the similar results with the F${}_{\text{𝑆𝑇}}$ values. The closest D${}_{\text{𝐴}}$ distance was observed between Chinese Yunnan Hani group and KHV (0.0032), followed by the HNH (0.0032) and CHS (0.0033); while the farthest D${}_{\text{𝐴}}$ distance was between Chinese Yunnan Hani group and YRI (0.1014), followed by ESN (0.0999) and MSL (0.0967). Two heatmaps constructed based on the values of pairwise F${}_{\text{𝑆𝑇}}$ and D${}_{\text{𝐴}}$ distances are exhibited through different colors in Fig. 4A,C, respectively. As shown in the heatmaps, Chinese Yunnan Hani group had larger F${}_{\text{𝑆𝑇}}$ values and D${}_{\text{𝐴}}$ distances with the African populations but smaller F${}_{\text{𝑆𝑇}}$ values and D${}_{\text{𝐴}}$ distances with reference East Asian populations. The pairwise F${}_{\text{𝑆𝑇}}$ values and D${}_{\text{𝐴}}$ distances between Chinese Yunnan Hani group and 30 reference populations were also displayed in the line graphs in Fig. 4B,D, respectively.

Fig. 4.

The heatmaps and the line graphs of the pairwise F${}_{S\mathit{}T}$values and D${}_A$ distances. (A) The heatmap of pairwise F${}_{\text{𝑆𝑇}}$values among Chinese Yunnan Hani group and 30 reference populations. (B) The line graph of the pairwise F${}_{\text{𝑆𝑇}}$ values between Chinese Yunnan Hani group and 30 reference populations. (C) The heatmap of the pairwise D${}_{\text{𝐴}}$ distances among Chinese Yunnan Hani group and 30 reference populations. (D) The line graph of the pairwise D${}_{\text{𝐴}}$ distances between Chinese Yunnan Hani group and 30 reference populations.

The pairwise F${}_{\text{𝑆𝑇}}$ values of five intercontinental populations in pairs, representing the same intercontinental populations as a whole, were also calculated, and the results are presented in Supplementary Table 9 and visualized through a boxplot in Fig. 5, respectively. The rs3076465, rs71852971, rs59841142, rs3067397, rs72031009, and rs34419736 loci exhibited relatively high F${}_{\text{𝑆𝑇}}$ values (F${}_{\text{𝑆𝑇}}$ $>$0.15) between African populations and other intercontinental populations. The rs145577149, rs113011930, rs145010051, rs76158822, and rs77635204 loci showed relatively high F${}_{\text{𝑆𝑇}}$ values (F${}_{\text{𝑆𝑇}}$ $>$0.15) between East Asian populations and other intercontinental populations, implying that these loci might be informative for discriminating East Asian populations from other intercontinental populations.

Fig. 5.

The pairwise F${}_{S\mathit{}T}$ values at 57 autosomal InDels among pairwise intercontinental populations.

Two phylogenetic trees, one rooted tree in Fig. 6A and one unrooted tree in Fig. 6B, were also used to assess the genetic relationships of the 31 populations. As depicted in Fig. 6A, the loop phylogenetic tree created using the pairwise D${}_{\text{𝐴}}$ distances was mostly separated into two branches: seven populations from Africa were classified into the first branch, and Chinese Yunnan Hani group and other populations made up the second branch. In the second branch, four populations from Europe formed into one subbranch while the rest populations grouped into another subbranch. Chinese Yunnan Hani group clustered with nine reference East Asian populations. These trees indicated that the genetic relationships among Chinese Yunnan Hani group and reference East Asian populations were closer than those of the other intercontinental populations. A similar branch distribution pattern could be observed in the unrooted tree in Fig. 6B, which was derived from the allelic frequencies of the 57 autosomal InDels in the 31 worldwide populations.

Fig. 6.

The phylogenetic relationship reconstructions among Chinese Yunnan Hani group and 30 reference populations. (A) The phylogenetic tree conducted based on the pairwise D${}_{\text{𝐴}}$ distances among Chinese Yunnan Hani group and 30 reference populations. (B) The unrooted tree developed based on the allele frequencies of 57 autosomal InDels among Chinese Yunnan Hani group and 30 reference populations.

The PCA and MDS analyses were also employed to further investigate the genetic relationships among Chinese Yunnan Hani group and 30 worldwide reference populations, which are exhibited in Fig. 7. The results of genotype-based PCA plots at the individual level are exhibited in Fig. 7A–D, while the results of allele frequency-based PCA chart at the population level is presented in Fig. 7E. The African, reference East Asian, and European populations was in Fig. 7A, while South Asian populations, American populations, and Chinese Yunnan Hani group are shown in Fig. 7B–D, respectively. The individuals from African, European and reference East Asian populations could be clearly distinguished. Most of Chinese Yunnan Hani individuals overlapped with reference East Asian populations. At the population level in Fig. 7E, the first two principal components contributed cumulatively 69.50% of the total variation, with PC1 and PC2 accounting for 46.29%, and 23.21%, respectively. PC1 could categorize African, reference East Asian populations, and Chinese Yunnan Hani group from the rest of the populations; PC2 mainly separated African, European populations from other populations. The cos2 values of 57 autosomal InDels are shown in Fig. 8, and the rs112879447, rs10626599, rs61490765, rs151335218, and rs10590825 loci are located in the relatively inner part of the ring and presented relatively low cos2 values. The result of the MDS analysis in terms of pairwise F${}_{\text{𝑆𝑇}}$ values is shown in Fig. 7F. A similar pattern of population distributions was observed in the MDS plot, which further confirmed the close relationships between Chinese Yunnan Hani group and reference East Asian populations involved in this study.

Fig. 7.

The results of PCA and MDS analyses among Chinese Yunnan Hani group and reference populations. (A) The PCA plot on the individual level in 21 reference populations from three continents. (B) The PCA plot on the individual level in 26 reference populations from four continents. (C) The PCA plot on the individual level in 30 reference populations from five continents. (D) The PCA plot on the individual level among Chinese Yunnan Hani group and 30 reference populations. (E) The PCA plot on the population level among Chinese Yunnan Hani group and 30 reference populations. (F) The MDS analysis plot based on the pairwise F${}_{\text{𝑆𝑇}}$ values among Chinese Yunnan Hani group and 30 reference populations.

Fig. 8.

The cos2 values on the basis of allelic frequencies of 57 autosomal InDels in Chinese Yunnan Hani group and 30 reference populations.

To further demonstrate the ancestral compositions of Chinese Yunnan Hani group, population genetic structure analyses were also performed, and the results are presented in Fig. 9 and Supplementary Figs. 1,2, respectively. As shown in Supplementary Fig. 1, the Delta K maximize was observed at K = 3, which meant the optimum K value estimated by the Structure Harvester Website was three. Supplementary Fig. 2 showed the locations of the 31 global reference populations on the world map and the structure analysis results (K = 3) at the 31 population level through the pie plots. The estimated ancestral proportions of Chinese Yunnan Hani group and 30 reference populations by assuming three ancestral populations are shown in the boxplot in Fig. 9A. The structure analysis results (K = 3) at 4229 individual levels are exhibited in the circle plot in Fig. 9B and the triangle plot in Fig. 9C, respectively. The dots representing Chinese Yunnan Hani individuals overlapped for the most part with those representing the reference East Asian populations in Fig. 9C. Moreover, the detailed ancestral proportion (K = 3) of Chinese Yunnan Hani group and other 30 reference populations were also presented through the circle chart in Fig. 9D. When K = 3, three ancestry compositions of the African, East Asian, and European clusters were further divided, and the dominant ancestry component in Chinese Yunnan Hani group was East Asian. More specifically, the percentages of African, European, and East Asian ancestry components in Chinese Yunnan Hani group were 0.0146, 0.0455, and 0.9399, respectively. Ancestral compositions of Chinese Yunnan Hani group were always in accord with other East Asian populations, indicating that the genetic architecture between Chinese Yunnan Hani group and East Asian populations was more similar than those of other intercontinental populations.

Fig. 9.

The outcomes of population genetic structure analyses between Chinese Yunnan Hani group and 30 reference populations. (A) Estimated ancestry proportions of Chinese Yunnan Hani group and 30 reference populations by assuming three ancestral populations. (B) The population genetic structure analysis results (K = 3) on the individual level among Chinese Yunnan Hani group and 30 reference populations. (C) Clustering analysis results (K = 3) for individual ancestry estimation among Chinese Yunnan Hani group and 30 reference populations. (D) The population genetic structure analysis results (K = 3) on the population level among Chinese Yunnan Hani group and 30 reference populations.