We used a 9948 DNA (1 ng/μl) sample to evaluate PCR amplification conditions, accuracy, DNA mixture deconvolution, and inhibitor tolerance. Further, a sample containing 9947A DNA was used to assess the male specificity of the HomyGene RM Y32 kit. DNA samples (1 ng/μl) from two unrelated male individuals were used to perform the mixture analysis. Common bacterial and animal species, including colibacillus, chicken, pig, dog, sheep, cat, and cow, were chosen to evaluate the species specificity of the kit. Mock forensic case-type samples including hair root, blood stain (wall), saliva stain (table top), semen stain (underpants), and cell phone swabs were obtained from seven anonymous males. For the hair root samples, we obtained three hair roots from each individual. In addition, we collected samples from 232 unrelated healthy Han individuals living in the Guizhou province, China. All individuals in this study provided their written informed consent for participation. This research conformed to the guidelines and obtained the approval of the ethics committee of Guizhou Medical University.

RM Y-STRs that showed high mutation rates or genetic diversities were chose from previous studies (Huang et al., 2005; Lin et al., 2006; Ballantyne et al., 2010, 2012; Zhang et al., 2014; Meng et al., 2019; Liu et al., 2020a). Next, the primer pair for each RM Y-STR locus was designed using the Primer 3 online tool (Untergasser et al., 2012) and was synthesized by the Sangon Biotech (Shanghai, China).

All experiments were conducted according to the following parameters unless stated otherwise. The PCR cocktail consisted of 10 μl of reaction mix (HomyGene), 5 μl of RM Y32 primer mix (HomyGene), 1 μl of C-Taq polymerase (HomyGene), 1 μl of DNA sample (1 ng/μl), and 8 μl of sdH₂O. Multiplex PCR of each sample was performed on the ProFlex PCR system (Thermo Fisher Scientific). The detailed reaction conditions were as follows: 95°C for 2 min, 10 cycles of 94°C for 30 s, 58.5°C for 60 s, and 72°C for 60 s; 20 cycles of 90°C for 30 s, 57°C for 60 s, and 72°C for 60 s; and 72°C for 10 min.

PCR product (1 μl volume) was added to the mixture containing 10 μl of deionized HiDi Formamide (Thermo Fisher Scientific) and 0.3 μl of AGCU Marker SIZ-500 (AGCU ScienTech Incorporation, Wuxi, Jiangsu, China). We performed capillary electrophoresis using the 3500xL Genetic Analyzer (Thermo Fisher Scientific) equipped with 36-cm capillary arrays with POP-4^® Polymer (Thermo Fisher Scientific). First, spectral calibration was conducted on the 3500xL Genetic Analyzer using the HomyGene Six Dye Matrix Standard kit (HomyGene). Next, we separated and detected amplified products of 32 RM Y-STRs on the 3500xL Genetic Analyzer. GeneMapper^® ID-X Software v1.5 (Thermo Fisher Scientific) was used to assess allele typing of each locus by comparison with the allelic ladder. The 150 relative fluorescence unit (RFU) was used as the analytical threshold to detect allele peak, unless stated otherwise.

The allelic ladder for this assay was constructed according to the previous study (Jia et al., 2021). In summary, a series of DNA samples, which showed different allelic variations of 32 RM Y-STRs in population, were sequentially amplified using the designed primer pairs. For each RM Y-STR locus, the amplified product of each allele was collected, purified, and mixed to obtain an allelic ladder of the locus. Then, the allelic ladder of each RM Y-STR locus was mixed in reasonable proportions to produce the allelic ladder for the kit.

We used a 1 ng 9948 DNA sample to perform PCR in the presence of different concentrations of amplification reagents or various reaction conditions to determine the optimal reaction parameters of the kit. Only the index to be tested was changed, and other parameters remained fixed. Each test was conducted in triplicate. The testing conditions are listed as below (the parameter in bold was the recommended condition):

Annealing temperatures: 57.5°C and 56°C, 58°C and 56.5°C, 58.5°C and 57°C, 59°C and 57.5°C, and 59.5°C and 58°C.

PCR Cycles: 28, 30, and 32 cycles

The 1 ng 9948 DNA sample was amplified and detected by three different operators to validate the genotyping accuracy of the kit. In addition, 10 bloodstain samples were randomly selected and subjected to multiplex PCR amplification and allelic typing using the AGCU Y37 kit (AGCU ScienTech Incorporation) to assess the genotype concordances across different kits. The allelic ladder was injected and tested on the 3500xL Genetic Analyzer 24 times to analyze the precision of the allele size.

Different bloodstain cards including FTA, Haoyuan, Kecaifeng reinforce, Kecaifeng mini, Dabo, Xinhai, Xinhai 503, filter paper, Bokun classic, Bokun reinforce, and Bokun mini cards were used to store blood samples from the same individual. The storage time of these bloodstain cards ranged from 3 to 5 years. These cards were assessed by the developed system to evaluate sample compatibility.

The DNA samples of mock case-type samples were extracted by the EZ1 DNA Investigator kit (Qiagen, Hilden, Germany). Next, these DNA samples were amplified by the developed system. Finally, amplified products were separated and typed as indicated above, respectively.

The 9948 and 9947A DNA samples were mixed in different ratios (19:1, 9:1, 3:1, 1:1, 1:3, 1:9, and 1:19) to evaluate the male specificity of the kit. Next, DNA samples from two unrelated males were mixed in different ratios (19:1, 9:1, 3:1, 1:1, 1:3, 1:9, and 1:19) to evaluate the power of the kit to discriminate the mixture. The mixtures were amplified in triplicate using the developed kit. For these mixtures, the 50 RFU was viewed as the analytical threshold to detect allele peak.

The control sample 9948 (1 ng/μl) was serially diluted into 1, 0.5, 0.25, 0.125, and 0.0625 ng. Next, these samples, in triplicate, were used to evaluate the detection limit of the kit.

A 1 ng DNA sample from different species was used to assess the cross-reaction of the kit. Each test was conducted three times.

Frequently encountered PCR inhibitors were selected to evaluate the stability of the kit. These inhibitors were heme (300, 400, 500, and 600 μM), hemoglobin (50, 100, 200, and 260 μM), humic acid (50, 150, 200, and 260 ng/μl), indigo (12, 14, 17, and 20 mM), Ca²⁺ (1.0, 1.5, 2.0, and 2.6 mM), and EDTA (1.0, 1.5, 2.0, and 2.5 mM). Each inhibitor was tested in triplicate.

Allelic frequencies and gene diversities (GDs) of 32 RM Y-STRs in Guizhou Han population were estimated using the STRAF online tool (Gouy and Zieger, 2017). Haplotype match probability (HMP), DC, and haplotype diversity (HD) of 32 RM Y-STRs in the Guizhou Han population were calculated on the basis of a previous study (Liu et al., 2020b).

The information relative to the 32 RM Y-STR loci is provided in Table 1. These 32 loci were classified into five sub-groups labeled by different dyes: DYS612, DYF387S1a/b, DYS627, and DYS518 (FAM); DYF404S1a/b, DYS534, DYS449, and DYS626 (HEX); DYS526a/b, DYS570, DYF399S1a/b/c, and DYS516 (TAMRA); DYS576, DYF403S1a1/a2/a3, DYF403S1b, and DYS547 (ROX); Y-GATA-A10, DYS458, DYS630, DYS464a/b/c/d, DYS446, and DYS713 (VIG). The amplicon ranges of these 32 RM Y-STRs are illustrated in Supplementary Figure S1. The amplicon lengths of these RM Y-STRs in the developed kit ranged from 90 to 500 base pairs (bp). The allelic profile of the 9948 DNA sample for these 32 RM Y-STRs is shown in Figure 1.

For these 32 RM Y-STRs, more than half of Y-STRs were those RM Y-STRs (DYF387S1, DYF399S1, DYF403S1, DYF404S1, DYS449, DYS518, DYS526, DYS547, DYS570, DYS576, DYS612, DYS626, and DYS627) identified by Ballantyne et al. (2010). Subsequently, the same research group also evaluated differentiation efficiencies of these RM Y-STRs in different continental and endogamous populations, and they found that these loci showed a higher male DC than the 17 Y-STRs from the Y filer kit (Thermo Fisher Scientific) (Ballantyne et al., 2012; Adnan et al., 2016). Zhang et al. assessed the mutation rates of these loci in the Hubei Han population and stated that their mutation rate ranged from 2.00 × 10^–3 to 4.59 × 10^–2 (Zhang et al., 2017). Yuan et al. (2019) also investigated the genetic distributions of these RM Y-STRs in the Beijing Han population, and they discerned similar results. For the remaining loci (DYS446, DYS458, DYS464, DYS516, DYS534, DYS630, DYS713, and Y-GATA-A10), Feng et al. (2020) found that DYS446, DYS458, DYS630, and Y-GATA-A10 displayed relatively high GD values (>0.5) in four minority groups from China. Zhou et al. (2021) discovered that mutation rates of DYS464 and DYS713 loci in the Sichuan Han population were 4 × 10^–3 and 2.4 × 10^–2, respectively. Fan et al. (2021b) reported that mutation rates of DYS516 and DYS534 loci in the coastal southeastern Han populations were 3.89 × 10^–3 and 1.07 × 10^–2, respectively. In summary, these 32 RM Y-STRs loci showed relatively high mutation rates and GD values across Chinese populations and could be beneficial in differentiating male lineages and unrelated males. However, we did not investigate the mutation rates of these 32 RM Y-STRs in the current study.

In practice, the annealing temperature may be affected by variations of the thermal cycler. Accordingly, it is essential to assess the effect of temperature variation on the Y-STR profile. For the HomyGene RM Y32 panel, two thermal cycle reactions are needed that possess different reaction conditions. Therefore, we simultaneously assessed the effect of annealing temperature 1 (first thermal cycle reaction) and annealing temperature 2 (second thermal cycle reaction) on the amplification performance of the developed kit. As shown in Supplementary Figure S2, full profile of 32 RM Y-STRs could be observed at different annealing temperatures. Whereas, some extra peaks were observed when the annealing temperature was 2°C and 1°C lower than the recommended annealing temperatures. Even so, these non-specific amplification products exerted little influence on allele detection. It should be noted that these artificial peaks may bring about some difficulties when evaluating mixture samples. From the above results, 58.5°C and 57°C were the optimal annealing temperatures for two different PCR cycles, respectively.

Forensic researchers often increase the number of cycles to improve detection rate of STRs in some trace samples. Here, we evaluated the effect of different cycle numbers on the Y-STR profile. As shown in Supplementary Figure S3, the allelic profile of 32 RM Y-STRs could be obtained at different cycle numbers. Besides, it was expected that peak height increased with increasing cycle numbers. Given that a better peak height balance was observed at 30 cycles, we proposed that the best cycle number was 30.

The concentration of PCR reagents may influence the multiplex amplification performance. It is critical to assess the effect of fluctuation of the PCR reagent concentrations on the overall appearance of Y-STRs. For the primer mix, all profile of 32 RM Y-STRs could be observed at various concentrations of primer mix (Supplementary Figure S4). Furthermore, no extra peaks were detected in the read region (90–490 bp), and the intra-color and inter-color balances were always greater than 70% and 50% at different concentrations of the primer mix, respectively. For reaction mix and C-Taq polymerase, similar phenomena were observed (Supplementary Figures S5, S6). Overall, the multiplex amplification system comprising 32 RM Y-STRs showed robust amplification and tolerated variations in PCR reagent concentrations.

In this study, three different researchers amplified and analyzed the 9948 DNA sample using the developed kit. Consistent genotype results for the 32 RM Y-STRs were obtained.

There are eight overlapping Y-STR loci between the system in this study and the AGCU Y37 kit. Therefore, 10 samples were randomly selected to evaluate the concordances of these RM Y-STRs. We found that overlapped RM Y-STR loci between two kits showed the same allelic profile for the same sample, implying that the kit exhibited good concordances.

Allele size accuracy is the indispensable prerequisite for obtaining a credible genetic profile. We injected an allelic ladder into the 24 capillaries on the 3500xL Genetic Analyzer to assess fluctuations in allele size. As shown in Figure 2, standard variations of each allele ranged from 0.0370 to 0.2400, implying a relatively small allele size variation. The results indicated that the developed kit achieved sufficient genotyping accuracy and size precision.

Forensic workers often use different bloodstain cards to store samples. It is of great importance to evaluate influence of different storage cards on the overall performance of the kit. Here, we amplified and analyzed bloodstain samples deposited in 11 types of commercially available storage cards (FTA, Haoyuan, Kecaifeng reinforce, Kecaifeng mini, Dabo, Xinhai, Xinhai 503, filter paper, Bokun classic, Bokun reinforce, and Bokun mini). We found that samples on bloodstain cards from the same individual showed a consistent allelic profile of 32 RM Y-STRs, indicating the kit possessed good sample adaptability.

We also assessed the power of the kit to detect case-type samples. The complete allelic profile of 32 RM Y-STR loci was obtained from simulated samples (hair root, blood stain on walls, saliva stains from table top, semen stains from underpants, and cell phone swabs). The allelic profile of a hair root sample from one individual for the 32 RM Y-STRs is shown in Supplementary Figure S7. More importantly, biological samples from the same donor produced identical allelic profiles for the 32 RM Y-STRs. Therefore, we proposed that the developed system could be utilized to detect common case-type samples.

Mixed samples are frequently encountered in sexual assault cases. Further, mixed samples also result from occasional contaminations from operator handling or from the workplace. Therefore, it is necessary to evaluate the power of the kit to discriminate DNA mixtures. Firstly, we mixed 9948 and 9947A samples at 19:1, 9:1, 3:1, 1:1, 1:3, 1:9, and 1:19 ratios and amplified these mixtures using the developed kit. As shown in Supplementary Figure S8, the full allelic profile of the male contributor was observed at 1:1, 1:3, 1:9, and 1:19 ratios. For male–male mixture, we constructed mixtures of two unrelated males in 19:1, 9:1, 3:1, 1:1, 1:3, 1:9, and 1:19 ratios. Similarly, a complete allelic profile of a minor contributor could be obtained in any mixed ratios (Supplementary Figure S9). In summary, the kit showed relatively a relatively high discrimination efficiency for mixed samples.

Biological samples collected from forensic scenes may contain trace amounts of DNA. Different quantities of DNA samples were tested to assess the detection limitation of the kit. As shown in Figure 3, full allelic profile of 32 RM Y-STRs could be observed at 1, 0.5, 0.25, 0.125, and 0.0625 ng. What is more, the peak height gradually decreased with the reduction of DNA amount. Therefore, the recommended quantity of DNA for the developed RM Y-STRs system is 1 ng.

Common animals sometimes exist in crime scenes. Thus, it is important to evaluate the species specificity of the kit. No cross-reactions of 32 RM Y-STRs were observed for seven common species tested (Supplementary Figure S10). Thus, we proposed that the kit exhibited an unexceptionable species specificity.

There are some PCR inhibitors that commonly exist in forensic biological samples. These inhibitors may interfere with the performance of multiplex amplification. In this study, we explored the tolerance of the kit to frequent inhibitors (heme, hemoglobin, humic acid, indigo, Ca²⁺, and EDTA). As shown in Figure 4 and Supplementary Figure S11, the complete allelic profile of 32 RM Y-STRs could be obtained even with the addition of 300 μM heme. When the concentration of heme increased to 400 and 500 μM, the peak height of larger amplicons decreased and most loci began ceased to be detected. Besides, no peaks were observed with 600 μM heme. With regard to PCR inhibition by hemoglobin, 32 RM Y-STRs showed complete allelic profile at 50 and 100 μM. When the concentration of hemoglobin increased to 200 μM, peak height began to decrease. Moreover, 3% of the 32 RM Y-STR loci were undetected at 260 μM hemoglobin (Figure 4 and Supplementary Figure S12). For humic acid, the large amplicons were missed at 150 ng/μl, and more alleles were undetected when humic acid increased to 200 and 260 ng/μl (Figure 4 and Supplementary Figure S13). With regard to indigo, the complete allelic profile of 32 RM Y-STRs were observed at 12 and 14 mM. However, when the indigo concentration increased to 17 mM, some large amplicons were undetected (Figure 4 and Supplementary Figure S14). For Ca²⁺, the allele peak height gradually decreased with increasing Ca²⁺ concentration (Figure 4 and Supplementary Figure S15). In addition, some alleles were undetected starting at 1.5 mM and only 78.26% of RM Y-STRs were observed at 2.6 mM. For EDTA, the full allelic profile was observed at 1.0 mM, whereas no profile was observed at 2.5 mM (Figure 4 and Supplementary Figure S16). Overall, the developed kit showed relatively good tolerance in the presence of these common inhibitors.

We investigated the genetic distributions and GD of 32 RM Y-STRs in the Guizhou Han population. The genetic profile and allelic frequencies of 32 RM Y-STRs in the Guizhou Han population are given in Supplementary Tables S1, S2, respectively. Furthermore, the number of alleles and GD values of 32 RM Y-STRs observed in the Guizhou Han population are also illustrated in Figure 5. The number of alleles for these 32 RM Y-STRs ranged from 6 (DYS516 and Y-GATA-A10) to 144 (DYF399S1a/b/c). The GD values of these RM Y-STRs distributed from 0.6828 (Y-GATA-A10) to 0.9942 (DYF399S1a/b/c). In a previous study, Chen et al. evaluated the forensic values of the 23 Y-STRs in the Guizhou Han population and found that four loci possessed low GD values (<0.5) (Chen et al., 2018). In the present study, the GD values of 32 RM Y-STRs were greater than 0.5, indicating that they exhibited relatively high genetic diversities in the GuiZhou Han population. What is more, no shared haplotype was observed for these 32 RM Y-STRs in 232 Guizhou Han individuals. The HMP, DC, and HD values of 32 RM Y-STRs in the Guizhou Han population were 0.0043, 1, and 1, respectively. Overall, these 32 RM Y-STRs showed high genetic polymorphisms in the Guizhou Han population and could be used to differentiate unrelated males.