We collected 312 bloodstain samples from unrelated healthy Guizhou Han individuals after obtaining their written informed consent. The 9948 and 9947A positive samples (1 ng/μL) were obtained from Promega Corporation (WI, United States ). DNA samples of common species including dog, pig, cow, sheep, chicken, mouse, rabbit, fish, and colibacillus were collected from the Animal Laboratory Center of Guizhou Medical University to assess species specificity of the STRtyper-27 comp kit. This research was performed in line with the guidelines of Guizhou Medical University and warranted by the Ethic Commission of Guizhou Medical University.

DNA sample of 1 ng was used to conduct the multiplex PCR of 27 loci according to the following specification. First, we prepared 10 μL PCR cocktail comprising 5 μL STRtyper-27 comp Master Mix, 2.5 μL STRtyper-27 comp Primer Mix, 2.5 μL ddH₂O, and 1 ng DNA sample. Second, PCR was conducted on the GeneAmp PCR System 9700 (Applied Biosystems, Foster City, CA, United States ) under reaction conditions of initial denaturation at 95°C for 5 min; 28 cycles of 94°C for 10s, 61°C for 60s, and 70°C for 30 s; and 60°C for 15 min. Third, we mixed 1 μL amplified product/STRtyper-27 comp Allelic Ladder Mix with 8.75 μL deionized HiDi Formamide and 0.25 μL ILS-500 (HEALTH Gene Technologies) and then denatured the mixture at 95°C for 3 min, followed by chilling at 4°C for 3 min. Finally, the mixture was electrophoresed and separated by the 3500xL Genetic Analyzer (Thermo Fisher Scientific). STR typing of each locus was determined by the GeneMapper® ID-X Software v1.5 (Thermo Fisher Scientific) in comparison with the allelic ladder.

Different annealing temperatures (56, 57, 58, 59, 60, 61, 62, 63, 64, 65, and 66°C), extension temperatures (65, 66, 67, 68, 69, 70, 71, 72, 73, 74, and 75°C), final extension time (5, 10, 15, 20, 25, and 30min), cycle numbers (25, 26, 27, 28, 29, and 30), and reaction volumes (5, 10, 15, 20, and 25 μL) were set to assess the performance of the kit in various PCR conditions. In addition, we also adjusted the concentrations of master mix (0.7×, 0.8×, 0.9×, 1.0×, 1.1×, 1.2×, and 1.3×) and primer mix (0.7×, 0.8×, 0.9×, 1.0×, 1.1×, 1.2×, and 1.3×) to explore the robustness of the kit for reagent fluctuations. The aforementioned experiments were performed by the 9948 DNA sample by adjusting the testing condition.

Cross-reaction of the kit with non-human samples were evaluated by amplifying DNA samples of dog, pig, cow, sheep, chicken, mouse, rabbit, fish, and colibacillus.

The 9948 DNA sample was serially diluted to explore the detection lower limit of the kit: 1ng, 500, 250, 125, and 62.5 pg/μL. In addition, we also assessed the detection upper limit of the kit: 1, 2, 5, and 10 ng/μL.

To evaluate size precision of the kit, we injected and detected the STRtyper-27 comp Allelic Ladder Mix by the 3500xL Genetic Analyzer 24 times.

Different ratios of 9948 and 9947A mixtures (1:1, 1:2, 1:4, 1:8, 1:19, 19:1, 8:1, 4:1, and 2:1) were constructed to assess the power of the kit to detect the mixture.

In total, seven inhibitors including heme (1.0, 1.2, 1.4, 1.6, and 1.8 mM), humic acid (0.4, 0.43, 0.45, and 0.47 mg), EDTA (1, 5, 10, 15, 20, and 25 mM), melanin (0.5, 0.6, 0.7, 0.8, and 0.9 mg), Ca²⁺ (16, 18, 20, 22, and 24 mM), and tannin (1, 3, 5, and 7 mg) were collected to evaluate the tolerance of the kit to these inhibitors.

We used the ultraviolet (wave length: 254nm; power: 28W) to treat the positive DNA sample 9948 (1 ng/μL) to simulate the degraded sample at different time periods (0, 15, 30, 45, and 60 min). Then, these mocked samples were detected by the developed kit in triplicate. Here, the 50-relative fluorescence unit was used as the detecting threshold to determine the allele peak.

Common samples found in the forensic scene including cigarette, bloodstain, seminal stain, fingerprint swab, and blood swab were collected and detected by the developed kit. First, we extracted DNA samples from these biomaterials by the ML ultrafine magnetic bead extraction kit (Changchun Bokun Biotech Corporation, Jilin, China). Next, these DNA samples were detected by the developed kit and the STRtyper-32G kit (HEALTH Gene Technologies).

Allelic frequencies and forensic-related parameters including expected heterogeneity (He), observed heterogeneity (Ho), polymorphism information content (PIC), match probability (PM), PD, power of exclusion (PE), and typical paternity index (TPI) of 26 STR loci in the Guizhou Han population were estimated by the STRAF online tool v1.0.5 (Gouy and Zieger, 2017). Furthermore, Hardy–Weinberg equilibrium (HWE) and linkage disequilibrium (LD) analysis of these STRs in the Guizhou Han population were also assessed by the STRAF online tool v1.0.5.

As shown in Figure 1, 27 loci of the STRtyper-27 comp kit were located on all chromosomes. These 27 loci were classified into four groups labeled by four dyes, respectively: D1S1656, D3S3045, D5S2800, D6S477, D9S1122, D13S325, and D18S1364 (FAM); D3S1744, D10S1435, D11S2368, D12S391, D21S2055, and D22S1045 (HEX); D4S2366, D6S474, D14S1434, D15S659, D16S539, D17S1301, and D22GATA198B05 (ROX); Amelogenin, D2S441, D7S3048, D8S1132, D10S1248, D19S253, and D20S482 (TAMRA). The allelic profile of 9947A positive sample is also given in Figure 2. The results showed that amplicon lengths of these loci distributed from 90 to 500 bp. Compared to other commercial STR kits (Hares, 2015; Zhu et al., 2015; Li et al., 2017; Ludeman et al., 2018; Zhong et al., 2019), we found that the kit in this study showed the most number (15) of overlapped loci using the Microreader 23SP kit (Supplementary Table S1). Even so, there were more than 10 novel STRs available in the developed kit. More importantly, the majority of loci presented in the kit were different from the expanded CODIS set. In addition, we found that physical distances between these novel STRs and those STRs (the expanded CODIS) on the same chromosomes were larger than 10 Mb (Supplementary Table S2), implying that these STRs could be viewed as independent loci from each other for forensic research. Even so, LD analyses of these STRs should be performed in the future. Anyway, we proposed that the kit in this study could be utilized as a high-efficient supplementary system for complex paternity analysis in parallel with the extant CODIS kits.

The annealing temperature is the key factor for PCR because it determines whether the primer binds to the DNA template. Therefore, annealing temperature variations may exert some effects on the performance of the multiplex detection assay. We assessed amplification efficiency of the developed kit at different annealing temperatures, as shown in Supplementary Figure S1. We found that the kit displayed comparable amplification performance at 56–64°C. However, some alleles begun to drop out at 65°C. In addition, allele peak height showed significant decrease at 65 and 66°C. Consequently, researchers are not suggested to set higher annealing temperature than our recommended temperature in practical application.

For the extension temperature, it is also crucial for PCR performance of the STR kit because it is related to the DNA template extension reaction. We evaluated the influences of different extension temperatures on the amplification performance of the developed kit, as presented in Supplementary Figure S2. The results revealed that all alleles could be correctly typed at different extension temperatures. Thus, extension temperature variations did not show any negative effects on amplification efficiency of the STR system.

The final elongation reaction can be used to avoid non-template–depended adenylation that may give rise to minus A or shoulder peaks. Different final elongation times at 60°C were set to evaluate amplification performance of the developed kit, as given in Supplementary Figure S3. The results demonstrated that all alleles showed normal electrophoretic peaks at different elongation times, indicating that the kit was tolerant to elongation time variations.

Different cycle numbers were tested to explore the optimal condition for the developed kit. As shown in Supplementary Figure S4, all alleles could be observed at different cycle numbers. In addition, allele peak height gradually increased with the augment of cycle numbers. Some non-specific amplification products were also observed at higher cycle numbers. Given that more balanced peak height was seen at 29 cycle numbers, we suggested that 29 is the optimal cycle number. Even so, researchers may explore the best condition for samples of interest in their studies.

Primer, Taq DNA polymerase, and PCR buffer are indispensable components in PCR. The fluctuations of PCR reagents may occur due to pipetting errors, which may have negative impacts on amplification efficiency. A series of concentrations of the primer mix and master mix comprising Taq DNA polymerase, PCR buffer, and other essential components were tested to evaluate the robustness of the developed kit. As shown in Supplementary Figure S5, a full allelic profile could be observed at different concentrations of the primer mix. Moreover, more balanced peak heights among different alleles were observed at 1.0× primer mix. For different concentrations of the master mix, we also found that all alleles could be detected (Supplementary Figure S6). However, some noise peaks were also observed at higher concentrations of the master mix. Thus, we recommended 1.0× master mix as the optimal concentration.

In forensic practice, researchers may reduce PCR reaction volume for trace samples. Thus, we also assessed the impact of different reaction volumes on the amplification performance of the developed kit. As given in Supplementary Figure S7, a full allelic profile could be obtained at different reaction volumes. In addition, allele peak height decreased with the increase of reaction volume. From the aforementioned results, we proposed that the developed kit is robust for different reaction volumes.

To determine detection limit of the developed kit, different quantities of DNA samples was amplified by the kit. Obtained results revealed that the complete allelic profile could be obtained from these diluted DNA samples except for 0.0625 ng DNA sample of which nearly 6% of alleles dropped out (Figure 3). In addition, as DNA quantity increased, allele peak height also gradually rose. Consequently, the developed kit is not recommended to detect those samples in which the amount of DNA was less than 0.0625 ng.

To evaluate the tolerance of the developed kit to common inhibitors, we added different concentrations of inhibitors to the PCR reagents. Obtained results are given in Figure 4; Supplementary Figures 8–13. For heme, we found all loci could be detected at 1.0 mM. In addition, alleles of some loci began to drop out at larger concentrations of heme, especially for 1.4–1.8 mM. For tannin, nearly 30% of 27 loci missed at 3mg, and more loci dropped out at 5–7 mg. For humic acid, we found that the majority of 27 loci could be observed at different concentrations of humic acid. Similar results could be seen from different concentrations of EDTA. For melanin and Ca²⁺, the loci detection rate of the kit decreased with the increasing of melanin and Ca²⁺ concentrations. Overall, we stated that the kit performed relatively good tolerances for these inhibitors.

Allele size variations may occur between different runs even on the same equipment, which affects reliable and accurate typing. Accordingly, it is vital to evaluate size precision of the kit. As shown in Figure 5, standard deviations of all alleles ranged from 0.02 to 0.08, indicating relatively subtle size variations of the kit. Thus, we proposed that the kit could provide accurate and reliable allelic typing.

DNA mixtures are common biological samples in forensic research. Consequently, it is necessary to assess the efficiency of the developed kit to detect mixtures. Two positive samples (9948 and 9947A) were mixed at different ratios (1:1, 1:2, 1:4, 1:8, 1:19, 19:1, 8:1, 4:1, and 2:1) and detected by the developed kit in triplicate. The allelic profile of 9948 and 9947A samples are presented in Supplementary Table S3. We found that nearly all alleles could be observed at different mixed ratios (Figure 6 and Supplementary Figure S14). Even so, one allele of D6S477 locus dropped out when mixed ratios were 1:4, 1:8, 1:19, 2:1, 8:1, and 19:1. One allele of D10S1435 locus was also missing at 1:19 ratio. Moreover, an extra allele was observed for D6S477 locus at 19:1 ratio. Given that the extra allele was less than one repeat unit than targeted alleles, we postulated that it might be stutter peaks. Anyway, we proposed that the developed kit could be employed to dissolve mixtures of two individuals given that alleles of most loci could be detected at different ratios.

Non-human DNA may present in the forensic scene. It is critical to evaluate species specificity of the developed kit. As shown in Supplementary Figure S15, no allele peaks were seen at nine common species including dog, pig, cow, sheep, chicken, mouse, rabbit, fish, and colibacillus, suggesting that the kit was human-specific and could be used to detect human samples without the interfering from other non-human samples.

Allele detection results of 27 loci for mocked degraded samples are presented in Table 1; Supplementary Figure S16. A complete allelic profile of these 27 loci could be obtained from degraded samples at 15 min exposure time. When the exposure time increased to 30min, more than 80% alleles of these 27 loci could be detected. In addition, more than 60% alleles of these 27 loci were observed at 45 min exposure time. More importantly, approximately 30% alleles of these 27 loci were still obtained even though the exposure time increased to 60 min, indicating the kit displayed relatively good tolerance to these mocked degraded samples.

For five collected samples (cigarette, bloodstain, seminal stain, fingerprint swab, and blood swab), we found that all alleles of 26 STRs could be observed. Furthermore, the same allelic profile of 14 overlapping STRs between the developed kit and the STRtyper-32G kit were also discerned from these samples. Thus, we proposed that the developed kit was suitable for detecting these common case samples.

First, we conducted HWE and LD tests of 26 STRs in the Guizhou Han population, as given in Supplementary Tables 4–5. The results demonstrated that these loci did not deviate from HWE in the Guizhou Han population (p < 0.05). For LD of pairwise loci (Supplementary Table S5), we found that nine pairs deviated from linkage equilibrium after applying Bonferroni correction (p < 0.000154). There are some factors that may lead to LD of pairwise loci, like genetic linkage, mutation, genetic drift, and population inbreeding (Slatkin, 2008). However, these nine pairs that deviated from linkage equilibrium in the current study were located on different chromosomes. It was less likely that genetic linkage resulted in LD of these loci. Accordingly, it might be other reasons like mutation and genetic drift leading to LD of these nine pairwise loci, which needed to be further evaluated.

Next, we investigated allelic distributions and forensic parameters of 26 STRs in the Guizhou Han population, as shown in Figure 7; Supplementary Table S4. A total of 259 alleles were observed at these loci with the number of alleles per locus ranging from 6 (D6S474) to 21 (D21S2055). For forensic parameters, we found that these 26 loci displayed small PM values less than 0.15; their PE, PD, and TPI values ranged from 0.3740 to 0.7771, 0.8599 to 0.9727, and 1.4857 to 4.5882, respectively. For diversity indexes, the mean Ho, He, and PIC values were 0.7839, 0.7872, and 0.7549. In addition, PIC values of all loci were larger than 0.6, indicating that these loci possessed high genetic diversities in the Guizhou Han population. The cumulative PD and PE values of 26 STRs in the Guizhou Han population were 0.999 999 999 999 999 999 999 999 999 996961 and 0.999 999 999 90297, respectively. Compared to other commercial kits (Hares, 2015; Zhu et al., 2015; Li et al., 2017; Ludeman et al., 2018; Zhong et al., 2019), the developed kit could obtain better cumulative PD and PE values, implying that the kit could be viewed as a high-performance system for forensic identity testing and paternity analyses in the Guizhou Han population. Of note, most loci out of the kit were different from the expanded CODIS. Therefore, not only did these 26 STRs enhance discrimination efficiency for unrelated individuals by combining with the available commonly used STRs, but they could also be used to assess complex kinships.