Skip to main content
NCBI home page
International Journal of Molecular Sciences logoLink to International Journal of Molecular Sciences
. 2026 Aug 8;27(16):7121. doi: 10.3390/ijms27167121

The Genetic Structure of the Romanian Male Population Revealed by 27 Y-Chromosome STR Loci

George Popoiu 1,2,*, Florin Stanciu 2,*, Paulina Popescu 2, Veronica Cuțăr 2, Simona Vladu 2, Violeta Nicola 2, Anastasia Procopciuc 2, Ana Rădulescu 2, Adnana Cotolea 2, Sergiu Emil Georgescu 1
Editor: Mariano Guardado-Estrada
PMCID: PMC13513713  PMID: 42653125

Abstract

Y-chromosomal Short Tandem Repeats (Y-STRs) provide insight into forensic evidence interpretation and male population history. Romania’s position at the crossroads of Europe and Western Asia makes it a key reference population for both domains, yet existing Romanian Y-STR databases are limited in geographic coverage and marker number. We genotyped 928 unrelated Romanian males across all 41 counties and Bucharest at 27 Y-STR loci using the Investigator Argus Y-28 Kit, and modeled haplotype frequencies using the Discrete Laplace (DL) distribution (disclapmix v1.7.5, 17-locus panel; disclapmix2 v0.6.1, 27-locus panel). Discrete Laplace modeling revealed 12 and 14 optimal mixture components for the 17- and 27-locus models, respectively, with median match probabilities of 1 in 1.25 × 107 and 1 in 5.8 × 1012, reflecting nearly six order-of-magnitude gain in evidential weight. Haplogroup analysis identified 16 Y-lineages; R1a, I2a, E1b1b, J2a, and R1b collectively represent ~80% of the population, and reflect Romania’s layered Steppe, Balkan, and Near Eastern ancestry, with subtle regional gradients between Moldavia, Transylvania, and Muntenia. This dataset provides validated forensic reference data and new insights into the patrilineal genetic structure of the Romanian population.

Keywords: Y-STR, Romania, haplotype diversity, forensic genetics, Discrete Laplace, haplogroup, match probability (PM), population genetics

1. Introduction

Forensic genetics is a branch of forensic science that applies molecular biology principles and methods to generate and interpret genetic evidence from biological samples recovered in the course of criminal investigations, providing information such as identity, biological relationships, or the possible origin of a trace. The results must be reported and interpreted within the bounds of what the underlying data can support.

The most widely used forensic genetics analyses are those that highlight a series of autosomal Short Tandem Repeat (STR) markers because they generate a genetic profile that is theoretically unique for each individual (except for monozygotic twins).

Even though this type of analysis offers numerous advantages and continues to be the gold standard in forensic genetics, there are also situations in which its efficacy is often challenged in forensic scenarios involving complex mixtures—most notably in sexual assault investigations where a high concentration of female deoxyribonucleic acid (DNA) can mask a trace amount of male genetic material. In these situations, the use of Y-chromosomal Short Tandem Repeats (Y-STRs) allows for the selective amplification of male DNA [1].

The forensic utility of Y-STRs arises not only from their ability to provide genetic information in male-dominant cases but the unique inheritance pattern of the Y-chromosome—passed almost entirely unchanged from father to son—grants it a distinct role in genealogical research, familial searching, and paternal lineage analysis [2]. This patrilineal transmission allows forensic geneticists to link unidentified remains to paternal relatives across multiple generations or to trace geographical ancestry through Y-haplogroups. However, the lack of recombination presents a unique statistical challenge: because changes in the structure of these markers only occur in the case of mutational events, many of the males in a paternal lineage can share the same Y-STR haplotype.

On the other hand, Romania occupies a strategically important position at the crossroads of Europe and the Near East and reflects a complex history of population movements including Neolithic farmer migrations from Anatolia, Bronze Age expansions from the Eurasian steppe, and subsequent Slavic and other historic-era gene flow [3,4]. Also, the previous Y-STR studies on Romanian samples have been limited in geographic coverage or marker number [5,6,7,8,9,10].

In order to better assess the weight of evidence obtained as a result of Y-STR typing, we conducted a study of the Romanian population to determine the range of the frequencies of Y-STR haplotypes using the Discrete Laplace distribution model, the Y-chromosome haplogroup frequencies and the regional particularities among the three historical regions, Moldavia, Transylvania and Muntenia.

Prior Y-STR investigations in Romania were conducted on smaller, geographically restricted samples and with marker panels that predate the extended commercial kits currently in routine forensic use. The present study is the first to combine complete national geographic coverage—sampling from all 41 Romanian counties and the municipality of Bucharest—with the full 27-locus profile generated by the Investigator Argus Y-28 Kit.

2. Results

2.1. Data Characteristics

All 928 haplotypes were unique when all 27 Y-STR markers were used. A summary of the analyzed data can be found in Table 1.

Table 1.

Number of alleles per locus, gene diversity (GD), polymorphism information content (PIC), match probability (PM) using the counting method and power of discrimination (PD) for 27 Y-STR markers.

PD PM (Counting Method) PIC GD Highest Allele Found Lowest Allele Found No. of Unique Alleles Locus
0.770 0.230 0.732 0.771 18 12 7 DYS19
0.831 0.169 0.810 0.832 18 9 11 DYS385a
0.802 0.198 0.776 0.803 21 10 13 DYS385b
0.509 0.491 0.458 0.509 16 10 6 DYS389I
0.754 0.246 0.714 0.755 34 26 9 DYS389II
0.724 0.276 0.680 0.725 27 19 9 DYS390
0.517 0.483 0.426 0.518 13 8 6 DYS391
0.350 0.650 0.327 0.351 16 10 7 DYS392
0.518 0.482 0.461 0.519 15 10 6 DYS393
0.586 0.414 0.507 0.586 16 13 5 DYS437
0.696 0.304 0.649 0.697 13 8 6 DYS438
0.714 0.286 0.663 0.714 15 9 7 DYS439
0.593 0.407 0.534 0.594 23 15 10 DYS448
0.834 0.166 0.814 0.835 39 25 13 DYS449
0.708 0.292 0.667 0.709 19 12 8 DYS456
0.773 0.227 0.737 0.774 22.2 13 15 DYS458
0.657 0.343 0.590 0.657 14 8 6 DYS460
0.845 0.155 0.829 0.846 32 18 15 DYS481
0.834 0.166 0.813 0.835 45 34 12 DYS518
0.620 0.380 0.570 0.621 15 8 8 DYS533
0.632 0.368 0.580 0.632 15 7 7 DYS549
0.806 0.194 0.779 0.807 25 12 12 DYS570
0.765 0.235 0.729 0.765 25 13 11 DYS576
0.864 0.136 0.850 0.865 24 14 11 DYS627
0.737 0.263 0.700 0.738 27 19 9 DYS635
0.708 0.292 0.672 0.709 14 7 8 DYS643
0.620 0.380 0.544 0.621 14 9 6 YGATAH4
Open in a new tab

The number of alleles for each marker analyzed (Figure 1) varies from 5 (for DYS437) to 15 (for DYS458 and DYS481) with a mean value of 9 alleles per locus and a total of 243 alleles found in the 928 samples analyzed.

Figure 1.

Figure 1

Open in a new tab

Distribution of Y-STR markers based on the number of unique alleles found.

As expected for haploid markers the values for power of discrimination (Figure 2) and gene diversity are similar, with the lowest values found at DYS392 marker (0.350 for PD and 0.351 for GD) and the highest at DYS627 (0.864 for PD and 0.865 for GD).

Figure 2.

Figure 2

Open in a new tab

Distribution of Y-STR markers based on PD values.

The data obtained also revealed that of the 27 Y-STR markers analyzed, 23 presented just full alleles and only four of them showed microvariants. In total, nine alleles with microvariants were observed, one each for the loci DYS390 (24.3), DYS437 (13.1), DYS448 (16.4) and six at the DYS458 locus (17.2, 18.2, 19.2, 20.2, 21.2, and 22.2). All nine designations are listed among previously cataloged alleles for these loci in the Y-Chromosome Haplotype Reference Database (YHRD, https://yhrd.org), and none represent a novel or previously unreported microvariant.

2.2. Discrete Laplace Analysis

2.2.1. Using Disclapmix v1.7.5 [11,12]

Although all 928 haplotypes are unique when the full 27-locus profile is considered, the R package disclapmix v1.7.5 only supports haplotype frequency modeling for a defined 17-locus subset (DYS19, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, DYS437, DYS438, DYS439, DYS448, DYS456, DYS458, DYS635, YGATAH4, DYS385a and DYS385b). When the 928 haplotypes are considered at these 17 loci only, 899 distinct profiles are observed (872 singletons, 25 haplotypes shared by two individuals, and 2 haplotypes shared by three individuals), reflecting the reduced discriminatory power of the smaller marker panel rather than any exclusion of samples.

Mixture model analysis of the haplotypes using the Discrete Laplace distribution performed with disclapmix v1.7.5 yielded Bayesian Information Criterion (BIC) values that decreased from 47,293 at k = 1 to a global minimum of 34,192 at k = 12 clusters, before rising slightly and fluctuating for k > 12. The 12-cluster solution was therefore adopted as optimal. The BIC profile showed the steepest descent between k = 1 and k = 6, indicating that the first six clusters capture the bulk of haplotypic structure in the Romanian population. The predicted haplotype frequencies for the 928 haplotypes analyzed varied between 2.58 × 10−18 and 4.97 × 10−3, with the median value at around 8.01 × 10−8 (Figure 3).

Figure 3.

Figure 3

Open in a new tab

Haplotype count and percent per frequency cluster using disclapmix v1.7.5 (percentages are rounded to two decimal places and do not sum to exactly 100% due to rounding).

The match probabilities computed for the 928 typed haplotypes varied between 1 in 2.01 × 102 and 1 in 3.88 × 1017 with the median estimated match probability of approximately 1 in 1.25 × 107.

2.2.2. Using Disclapmix2 v0.6.1 [13]

Mixture model analysis using disclapmix2 v0.6.1 resulted in BIC values that decreased from 81,469 at k = 1 to a global minimum of 62,320 at k = 14 clusters, before rising slightly for k > 14. In the end, the 14-cluster solution was adopted as optimal.

The predicted haplotype frequencies for the 928 haplotypes analyzed varied between 4.31 × 10−28 and 2.86 × 10−6, with the median value at around 1.73 × 10−13 (Figure 4).

Figure 4.

Figure 4

Open in a new tab

Haplotype count and percent per frequency cluster using disclapmix2 v0.6.1.

For the disclapmix2 v0.6.1, the match probabilities values ranged from 1 in 3.50 × 105 to 1 in 2.32 × 1027 with the median estimated match probability of approximately 1 in 5.8 × 1012.

The observed haplotype diversity was the same with disclapmix v1.7.5 and disclapmix2 v0.6.1 of 0.9989, but the model-based haplotype diversity differed between the two software packages, disclapmix v1.7.5 generating a value of 0.9272 while disclapmix2 v0.6.1 generated a value of 0.8292.

2.3. Y-Chromosome Haplogroup Diversity (HGD) Within the Romanian Population

Haplogroup assignment was successful for all 928 samples; 16 distinct haplogroups were identified. Supplementary Table S4 lists, for each of the 928 samples, the predicted haplogroup, the fitness score, and the Bayesian probability returned by Athey’s Haplogroup Predictor.

Table 2 presents, for the 16 distinct haplogroups identified, their absolute and relative frequencies for the total Romanian sample and the three regional subgroups.

Table 2.

Y-chromosome haplogroup counts and relative frequencies (%) in the total Romanian sample and three historical regional subgroups.

Muntenia (%) Muntenia (n) Transylvania (%) Transylvania (n) Moldavia (%) Moldavia (n) Romania (%) Romania (n) Y Haplogroup
0 0 0.32 1 0 0 0.11 1 E1b1a
17.78 72 16.23 50 9.30 20 15.30 142 E1b1b
2.72 11 2.27 7 4.19 9 2.91 27 G2a
5.68 23 7.14 22 4.65 10 5.93 55 I1
15.80 64 15.26 47 18.14 39 16.16 150 I2a
2.72 11 1.95 6 2.79 6 2.48 23 I2b
1.73 7 1.30 4 1.86 4 1.62 15 J1
17.78 72 12.34 38 14.42 31 15.19 141 J2a
3.70 15 3.57 11 3.26 7 3.56 33 J2b
0 0 0.32 1 0.47 1 0.22 2 L
0.99 4 1.30 4 2.79 6 1.51 14 N
0.25 1 0 0 0 0 0.11 1 Q
16.05 65 23.70 73 23.26 50 20.26 188 R1a
13.58 55 13.96 43 13.49 29 13.69 127 R1b
0.25 1 0 0 0.47 1 0.22 2 R2
0.99 4 0.32 1 0.93 2 0.75 7 T
100.00 * 405 100.00 * 308 100.00 * 215 100.00 * 928 Total
0.8610 ± 0.0248 0.8518 ± 0.0291 0.8584 ± 0.0338 0.8608 ± 0.0164 Haplogroup diversity
Open in a new tab

* Percentages are rounded to two decimal places and do not sum to exactly 100% due to rounding.

At the national level, the five most frequent haplogroups were R1a (n = 188; 20.26%), I2a (n = 150; 16.16%), E1b1b (n = 142; 15.30%), J2a (n = 141; 15.19%), and R1b (n = 127; 13.69%), together accounting for 80.6% of the sample. This pentapartite haplogroup structure reflects the layered population history of Romania: R1a is the principal Steppe-ancestry lineage associated with Indo-European expansion [3,14]; I2a is the most dominant lineage in Balkan region which borders Romania to the south and with which there have been close historical ties [15]; E1b1b possibly represents migrations from North Africa and the Near East to Europe [16]; J2a reflects Anatolian and subsequent Near Eastern gene flow that are often associated with later Neolithic and Bronze Age movements into Europe [16,17]; and R1b which is associated with populations that migrated from the Eurasian Steppe into Europe like the Yamnaya culture [3,18]. Rarer haplogroups—including G2a (2.91%) [19], I1 (5.93%) [20], J2b (3.56%) [16], and N (1.51%) [21]—reflect additional Caucasian Neolithic, Northern European, Balkan-South Asian, and Uralic contributions, respectively.

Regional comparisons (Figure 5) revealed that R1a was jointly the most frequent haplogroup in Moldavia (23.26%) and Transylvania (23.70%), while J2a shared the top position in Muntenia (17.78%, tied with E1b1b). Moldavia showed the lowest E1b1b frequency (9.30%) compared with Transylvania (16.23%) and Muntenia (17.78%). I2a was most prevalent in Moldavia (18.14%). N was more frequent in Moldavia (2.79%) than in Muntenia (0.99%).

Figure 5.

Figure 5

Open in a new tab

Y-chromosome haplogroup distribution map by region (Map created using DiMarco, M. Datamaps: Customizable SVG map visualizations for the web).

To formally test whether these regional differences were statistically significant, an Analysis of Molecular Variance (AMOVA) was performed on the full 27-locus haplotypes using the RST estimator [22]. The global three-region AMOVA yielded RST = 0.0026, indicating that only 0.26% of the total molecular variance was attributable to differences among regions; this result was nonetheless statistically significant (10,000 permutations, p = 0.027). Pairwise comparisons showed that Moldavia and Muntenia differed significantly (RST = 0.0052, p = 0.009), whereas Moldavia versus Transylvania (RST = −0.0005, p = 0.527) and Transylvania versus Muntenia (RST = 0.0023, p = 0.062) did not reach significance. These haplotype-level results indicate that, although the Romanian male population is overall genetically homogeneous, a small but detectable degree of regional substructure exists, driven mainly by differentiation between the Moldavia and Muntenia subgroups.

Haplogroup diversity was high and similar across all three regions (Moldavia: HGD = 0.8584 ± 0.0338; Transylvania: HGD = 0.8518 ± 0.0291; Muntenia: HGD = 0.8610 ± 0.0248), with the national estimate at HGD = 0.8608 ± 0.0164, indicating that the haplogroup-level inter-regional differentiation is modest relative to within-region diversity.

Supplementary Table S1 is an Excel spreadsheet that contains the haplotypes used and the predicted frequencies with disclapmix v1.7.5 R package, and the match probabilities calculated based on the frequencies obtained.

Supplementary Table S2 is an Excel spreadsheet that contains the haplotypes used and the predicted frequencies with disclapmix2 v0.6.1 R package, and the match probabilities calculated based on the frequencies obtained.

Supplementary Table S3 is an Excel spreadsheet that contains the alleles observed and the count of them in the analyzed samples for all 27 Y-STR loci.

Supplementary Table S4 is an Excel spreadsheet that contains Athey’s haplogroup predictions, fitness score and probability of the haplogroup for all 928 samples analyzed.

Supplementary Figure S1 is an HTML file which features an interactive map showing the distribution of Y-STR based haplogroups by historical region.

Supplementary Figure S2 is a .tif file which features the haplotype percent per frequency cluster using disclapmix v1.7.5 in a pie chart format (percentages are rounded to two decimal places and do not sum to exactly 100% due to rounding).

Supplementary Figure S3 is a .tif file which features the haplotype percent per frequency cluster using disclapmix2 v0.6.1 in a pie chart format.

3. Discussion

The results demonstrate that using a larger number of Y-STR markers increases the chances of discrimination between individuals: using the 17 markers considered by disclapmix v1.7.5 out of the 928 haplotypes only 899 were unique, but when all 27 markers were considered as disclapmix2 v0.6.1 does, all 928 were unique.

Both methods correctly identify structured population substructure, but disclapmix v1.7.5 selects 12 clusters, while disclapmix2 v0.6.1 selects 14 clusters. The higher absolute BIC for disclapmix2 v0.6.1 is expected given that BIC scales with the number of parameters, and 27 loci with 14 clusters means 584 parameters versus ~243 for the disclapmix v1.7.5 model. The BIC improvement from k = 1 to the optimum is actually larger for disclapmix2 v0.6.1 (−19,149 vs. −13,101), indicating that the additional loci reveal more genuine genetic substructure.

The most consequential difference between the two methods used was observed in the predicted frequency values. The median predicted frequency shifts from 8 × 10−8 (disclapmix v1.7.5) to 1.7 × 10−13 (disclapmix2 v0.6.1)—nearly six orders of magnitude. In the disclapmix v1.7.5 analysis, 77.2% of haplotypes fall in the 10−9 to 10−3 frequency range, and no haplotypes fall below 10−18. In the disclapmix2 v0.6.1 analysis, 89.7% of haplotypes fall below 10−9, and 11.2% fall below 10−18—a range entirely inaccessible to the 17-locus model. This means that the forensic evidential weight of a haplotype match, expressed as the match probability, goes from “1 in 1.25 × 107” to “1 in 5.8 × 1012” unrelated males. For matched haplotypes, disclapmix v1.7.5 frequencies are consistently about 104–106 × higher than disclapmix2 v0.6.1 values, which is the direct mathematical consequence of conditioning on 10 fewer loci.

The wide power-of-discrimination range observed across the 27 markers—from 0.350 at DYS392 to 0.864 at DYS627—confirms that the panel’s evidential contribution is not dominated by a handful of highly polymorphic loci but distributed across the marker set. This redundancy provides resilience against partial profiles, which are common in degraded or low-template casework samples.

As we stated before, the highest gene diversity and power of discrimination were observed at DYS627 (GD = 0.865; PD = 0.864). The elevated diversity at DYS627 is consistent with its classification as a rapidly mutating Y-STR [23]. This pattern extends across the panel: the six RM Y-STRs included in this kit (DYS449, DYS481, DYS518, DYS570, DYS576, DYS627) rank among the ten most diverse of the 27 loci studied (GD = 0.826 ± 0.035 vs. 0.659 ± 0.117 for the remaining loci), consistent with their markedly higher mutation rate relative to conventional Y-STRs.

To further determine the discriminatory power of the 27 Y-STR markers, the Investigator Argus Y-28 Kit (Qiagen) should be tested on paternally related individuals, paying special attention to the rapidly mutating Y-STR markers that this kit contains (DYS449, DYS481, DYS570, DYS576, DYS518, and DYS627).

The diversity of haplogroups identified following data analysis reflects the background and Romania’s geographical position on the route of the great migrations at the crossroads between Western Asia and Central and Western Europe, as well as between Southern and Northern Europe. The slight gradient of R1a enrichment in Moldavia and Transylvania relative to Muntenia and the higher J2a frequency in Muntenia reported here is broadly consistent with previously proposed population-history narratives for South-Eastern Europe [17], though we emphasize that STR-based haplogroup prediction alone, without single nucleotide polymorphism (SNP) subclade resolution or direct haplotype comparison against neighboring populations, cannot itself test or confirm specific migration hypotheses.

So, because haplogroup prediction relied on STR-based data rather than SNP typing the results obtained only provide a general picture of the events that led to the current structure of the Romanian population. Future work integrating Y-chromosome Single Nucleotide Polymorphism (Y-SNP) genotyping should allow more precise sub-haplogroup resolution and refined intra-population structure analysis.

4. Materials and Methods

4.1. Sample Collection

In this study we used the same sample of 928 unrelated men, born in Romania, used in the population-based study to determine allele frequencies for autosomal STR markers published the previous year [24]. We proceeded in this way to allow future comparison of the particularities observed in the analysis of autosomal STR markers with those observed in the analysis of Y-chromosome STR markers.

The 928 samples were selected with the aim of reflecting the percentages of Romania’s population for each county and Bucharest, as reported in the 2021 census. For 33 counties and Bucharest, this target was met; for 3 counties, the number of samples collected came very close to the target (over 80% of the required samples were collected); and for 5 counties, the required number of samples was not reached, with the collection rate ranging from 14% to 57%. However, the results obtained are not affected because all the analyses in this study were conducted at the national or historical-region level, where sample sizes (n = 215–405) are sufficient to support reliable diversity and frequency estimates; county-level counts are reported in Table 3 to document geographic coverage rather than to support county-level inference.

Table 3.

Distribution of analyzed samples.

County Historical Region Number of Analyzed Biological Samples
Bacău Moldavia 32
Botoșani Moldavia 21
Galați Moldavia 26
Iași Moldavia 40
Neamț Moldavia 24
Suceava Moldavia 34
Vaslui Moldavia 20
Vrancea Moldavia 18
Alba ** Transylvania 7
Arad Transylvania 22
Bihor Transylvania 29
Bistrița-Năsăud Transylvania 16
Brașov Transylvania 29
Caraș-Severin Transylvania 13
Cluj Transylvania 36
Covasna Transylvania 10
Harghita * Transylvania 13
Hunedoara Transylvania 19
Maramureș Transylvania 24
Mureș Transylvania 27
Sălaj Transylvania 11
Satu-Mare ** Transylvania 4
Sibiu Transylvania 20
Timiș * Transylvania 28
Argeș Muntenia 30
Brăila Muntenia 15
Bucharest Muntenia 90
Buzău Muntenia 21
Călărași ** Muntenia 5
Constanța Muntenia 34
Dâmbovița Muntenia 25
Dolj * Muntenia 30
Giurgiu ** Muntenia 8
Gorj Muntenia 17
Ialomița Muntenia 13
Ilfov ** Muntenia 4
Mehedinți Muntenia 12
Olt Muntenia 20
Prahova Muntenia 36
Teleorman Muntenia 17
Tulcea Muntenia 10
Vâlcea Muntenia 18
Open in a new tab

* Counties with number of samples very close to the target (>80%); ** Counties with number of samples not reached (14–57%).

To verify the absence of close relatives among the 928 donors, a pairwise kinship analysis was performed using previously generated 21-locus autosomal STR profiles for the same sample set and population-specific allele frequencies [24]. Likelihood-ratio kinship tests (Paternity Index, Sibling Index) were computed for all 430,128 possible sample pairs. No parent–offspring pairs were detected among the 928 individuals. Two pairs showed statistically significant evidence of elevated relatedness after Bonferroni correction for multiple comparisons, and a third pair was borderline; however, cross-checking the Y-STR haplotypes of these candidates confirmed that none share a paternal lineage (haplotypes differed at 13–20 of 27 marker fields, well beyond the range expected for paternal relatives). As Y-STR haplotypes are inherited exclusively through the paternal line, this confirms that no donors in the present sample are related through the paternal lineage, and that the reported haplotype frequency and random match probability estimates are not affected by relatedness among donors.

This set of samples is represented by epithelial buccal cells collected from convicted persons, in accordance with national law number 76 of 2008, using EasiCollect® Plus (Qiagen, Venlo, The Netherlands) FTA card devices (n = 897) or buccal swabs (n = 31).

4.2. DNA Extraction

For most samples, polymerase chain reaction (PCR) amplification was performed directly from FTA card material, without a separate conventional DNA extraction step, but for some of the samples genomic DNA extraction was performed before Y-STR PCR amplification.

Genomic DNA was extracted with the QIAamp Investigator BioRobot Kit (Qiagen) on the BioRobot Universal System (Qiagen) when samples were collected using buccal swabs; and when direct PCR amplification from the FTA card material did not yield optimal results because the collection procedure was not performed properly and the amount of biological material recovered on FTA card was low, DNA extraction was performed using the QIAsymphony DNA Investigator Kit (Qiagen) on the QIAsymphony SP/AS instrument (Qiagen).

4.3. DNA Quantification

The quantification of DNA was performed only for the samples extracted with the QIAsymphony DNA Investigator Kit (Qiagen) on the QIAsymphony SP/AS instrument (Qiagen) using the Investigator Quantiplex Pro RGQ (Qiagen) on the RotorGene Q Real-Time System (Qiagen) in accordance with the kit manufacturer’s protocol.

4.4. Y-STR PCR Amplification

The PCR amplification was performed in accordance with the kit manufacturer’s protocol using the Investigator Argus Y-28 Kit (Qiagen) and a Veriti™ Thermal Cycler (ThermoFisher Scientific, Waltham, MA, USA) which targeted 27 Y-STR loci (DYS19, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, DYS437, DYS438, DYS439, DYS448, DYS456, DYS458, DYS635, YGATAH4, DYS385a, DYS385b, DYS449, DYS460, DYS481, DYS518, DYS533, DYS549, DYS570, DYS576, DYS627 and DYS643), and two quality control markers.

If the presence of microvariants was observed following genotyping, they were confirmed using the Yfiler™ Plus PCR Amplification Kit (ThermoFisher Scientific) and a Veriti™ Thermal Cycler (ThermoFisher Scientific) using the kit manufacturer’s protocol.

4.5. Detection and Genotyping

Following amplification, electrophoresis of PCR products was carried out using the ABI 3500 Genetic Analyzer (ThermoFisher Scientific) together with allelic ladders, positive and negative controls using the recommendations of the amplification kit manufacturer.

The data collected by the 3500 Genetic Analyzer (ThermoFisher Scientific) were analyzed and the genotypes were assigned using GeneMapper ID-X 1.6 as the amplification kit manufacturer’s protocol specified.

4.6. Statistical Analysis

The calculations for gene diversity (GD), polymorphism information content (PIC), match probability (PM) using the counting method and power of discrimination (PD) were performed using the online program STR Analysis for Forensics (STRAF) v2.2.2 [25].

The Discrete Laplace method was used for determining the mixture model clustering, haplotype frequencies and match probability [11,12]. The computations were performed with the R package disclapmix v1.7.5 for a 17 Y-STR marker subset (DYS19, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, DYS437, DYS438, DYS439, DYS448, DYS456, DYS458, DYS635, YGATAH4, DYS385a and DYS385b) and also with the R package disclapmix2 v0.6.1 using all 27 Y-STR markers analyzed [13]. All computations, including disclapmix v1.7.5 and disclapmix2 v0.6.1 analysis, were performed in R v4.5.3.

The duplicated locus DYS385 was represented differently between the two model implementations: for disclapmix v1.7.5 it was entered as two ordered pseudo-loci, DYS385a (lower repeat number) and DYS385b (higher repeat number), the standard workaround required because the original DL model [11,12] assumes a single integer-valued allele per locus, whereas for disclapmix2 v0.6.1, which was purpose-built to natively support multi-copy loci, partial repeats and null alleles [13], DYS385 was instead entered as a single multi-copy field. DYS389I and DYS389II were entered as separate loci in both analyses without adjusting for their known structural dependency (the DYS389II amplicon includes the DYS389I repeat region), following the convention used in the DL method’s own reference datasets [11,12], in which loci are modeled as conditionally independent given cluster membership so that correlated multi-locus patterns are captured through cluster assignment rather than through an explicit locus-pair covariance term. Regarding differential mutation rates, the DL model does not assume equal variability across loci: it estimates an independent dispersion parameter for every locus within every cluster, fitted directly from the observed within-cluster allele-count variation [11,12].

Y-chromosome haplogroup prediction was performed with the web-based tool Athey’s Haplogroup Predictor [26] using only 25 of the 27 Y-STR markers (DYS518 and DYS627 are not included in the list of 111 markers of this tool). The criterion for designating a haplogroup to a haplotype was the highest value of the Bayesian probability returned by the prediction tool. Haplogroup diversity was estimated using Nei’s formula applied to haplogroup frequencies [27].

To evaluate whether the observed differences in haplotype composition among the three major historical regions (Moldavia, Transylvania, and Muntenia) were statistically significant, an Analysis of Molecular Variance (AMOVA) was performed using the RST estimator, which is appropriate for microsatellite loci evolving under a stepwise mutation model [22]. All AMOVA/RST calculations were implemented in Python (NumPy v2.5.1, SciPy v1.18.0).

5. Conclusions

This study delivers a fully characterized 27-locus Y-STR database for 928 Romanian males, representative of all counties and the three major historical regions. The use of disclapmix2 v0.6.1 across the extended marker panel not only achieved complete haplotype uniqueness but yielded match probabilities approximately six orders of magnitude lower than the 17-locus model, providing a significantly stronger evidential framework for casework. The documented haplogroup composition—led by R1a, I2a, E1b1b, J2a, and R1b—is consistent with Romania’s position as a demographic palimpsest shaped by Steppe migrations, Balkan Neolithic lineages, and Near Eastern gene flow.

The present dataset constitutes the largest geographically complete Y-STR reference resource for Romania published to date. Future work should focus on SNP-based sub-haplogroup resolution and on testing the Investigator Argus Y-28 Kit on paternally related individuals to further characterize discriminatory power within patrilineal groups.

Acknowledgments

This study was conceived as a result of studies conducted within the Doctoral School of Biology program of the University of Bucharest. This paper was supported by Council for Doctoral Studies (CSUD), University of Bucharest. Also, we would like to acknowledge and thank Romică Potorac, National Forensic Institute, for his support for research activities within the institute, and all the Department of Forensic Genetics auxiliary staff, without whom the completion of this study would have been more difficult.

Abbreviations

The following abbreviations are used in this manuscript:

BIC Bayesian Information Criterion
DL Discrete Laplace
DNA Deoxyribonucleic Acid
EU European Union
GD Gene diversity
HGD Haplogroup diversity
PCR Polymerase Chain Reaction
PD Power of discrimination
PIC Polymorphism information content
PM Match probability
SNP Single Nucleotide Polymorphism
STR Short Tandem Repeat
STRAF STR Analysis for Forensics
Y-SNP Y-chromosome Single Nucleotide Polymorphism
Y-STR Y-chromosome short Tandem repeat
Open in a new tab

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/ijms27167121/s1.

ijms-27-07121-s001.zip (451.9KB, zip)

Author Contributions

Conceptualization, G.P. and S.E.G.; methodology, G.P. and F.S.; validation, G.P. and F.S.; formal analysis, F.S. and G.P.; investigation, P.P., V.C., S.V., V.N., A.P., A.R., A.C., F.S. and G.P.; data curation, F.S. and G.P.; writing—original draft preparation, G.P.; writing—review and editing, F.S., S.E.G., A.P. and P.P.; visualization, G.P. and S.E.G.; supervision, F.S.; project administration, G.P. and S.E.G. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

The study was conducted in accordance with the code of ethics and deontology of the National Forensic Institute, the protocol was approved by the Ethics Committee on 27 October 2021 and all provisions of Regulation (EU) 2016/679 on the protection of natural persons with regard to the processing of personal data and the free movement of such data were respected during the handling of genetic information.

Informed Consent Statement

Informed consent was waived due to the fact that biological samples were taken from convicted persons in accordance with the provisions of the national Law no. 76 from 2008 on the organization and functioning of the National Forensic Genetic Data System.

Data Availability Statement

The original contributions presented in this study are included in the article and Supplementary Materials. Further inquiries can be directed to the corresponding authors.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

This research received no external funding.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

References

  • 1.Kayser M., Sajantila A. Mutations at Y-STR Loci: Implications for Paternity Testing and Forensic Analysis. Forensic Sci. Int. 2001;118:116–121. doi: 10.1016/S0379-0738(00)00480-1. [DOI] [PubMed] [Google Scholar]
  • 2.Jobling M.A., Tyler-Smith C. The Human Y Chromosome: An Evolutionary Marker Comes of Age. Nat. Rev. Genet. 2003;4:598–612. doi: 10.1038/nrg1124. [DOI] [PubMed] [Google Scholar]
  • 3.Haak W., Lazaridis I., Patterson N., Rohland N., Mallick S., Llamas B., Brandt G., Nordenfelt S., Harney E., Stewardson K., et al. Massive Migration from the Steppe Was a Source for Indo-European Languages in Europe. Nature. 2015;522:207–211. doi: 10.1038/nature14317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hervella M., Rotea M., Izagirre N., Constantinescu M., Alonso S., Ioana M., Lazăr C., Ridiche F., Soficaru A.D., Netea M.G., et al. Ancient DNA from South-East Europe Reveals Different Events during Early and Middle Neolithic Influencing the European Genetic Heritage. PLoS ONE. 2015;10:e0128810. doi: 10.1371/journal.pone.0128810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Barbarii L.E., Rolf B., Dermengiu D. Y-Chromosomal STR Haplotypes in a Romanian Population Sample. Int. J. Leg. Med. 2003;117:312–315. doi: 10.1007/s00414-003-0397-0. [DOI] [PubMed] [Google Scholar]
  • 6.Egyed B., Füredi S., Padar Z. Population Genetic Study in Two Transylvanian Populations Using Forensically Informative Autosomal and Y-Chromosomal STR Markers. Forensic Sci. Int. 2006;164:257–265. doi: 10.1016/j.forsciint.2005.10.020. [DOI] [PubMed] [Google Scholar]
  • 7.Stanciu F., Cuţăr V., Pîrlea S., Stoian V., Stoian I.M., Sevastre O., Popescu O.R. Population Data for Y-Chromosome Haplotypes Defined by 17 STRs in South-East Romania. Leg. Med. 2010;12:259–264. doi: 10.1016/j.legalmed.2010.05.007. [DOI] [PubMed] [Google Scholar]
  • 8.Bembea M., Patocs A., Kozma K., Jurca C., Skrypnyk C. Y-Chromosome STR Haplotype Diversity in Three Ethnically Isolated Population from North-Western Romania. Forensic Sci. Int. Genet. 2011;5:e99–e100. doi: 10.1016/j.fsigen.2010.12.004. [DOI] [PubMed] [Google Scholar]
  • 9.Varzari A., Kharkov V., Nikitin A.G., Raicu F., Simonova K., Stephan W., Weiss E.H., Stepanov V. Paleo-Balkan and Slavic Contributions to the Genetic Pool of Moldavians: Insights from the Y Chromosome. PLoS ONE. 2013;8:e53731. doi: 10.1371/journal.pone.0053731. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Hodișan R., Zaha D.C., Jurca C.M., Irina S., Bembea M. Genetic Diversity Based on the Analysis of 27 Y-Short Tandem Repetition (STR) Loci in Two Populations in the Apuseni Mountains, Romania. Cureus. 2024;16:e62505. doi: 10.7759/cureus.62505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Andersen M.M., Eriksen P.S., Morling N. The Discrete Laplace Exponential Family and Estimation of Y-STR Haplotype Frequencies. J. Theor. Biol. 2013;329:39–51. doi: 10.1016/j.jtbi.2013.03.009. Correction in J. Theor. Biol. 2014, 350, 110. https://doi.org/10.1016/j.jtbi.2014.03.001. [DOI] [PubMed] [Google Scholar]
  • 12.Andersen M.M., Eriksen P.S., Morling N. A Gentle Introduction to the Discrete Laplace Method for Estimating Y-STR Haplotype Frequencies. arXiv. 20131304.2129 [Google Scholar]
  • 13.Kruijver M., Taylor D., Buckleton J. Extending the Discrete Laplace Method: Incorporating Multi-Copy Loci, Partial Repeats and Null Alleles. Forensic Sci. Int. Genet. 2023;65:102876. doi: 10.1016/j.fsigen.2023.102876. [DOI] [PubMed] [Google Scholar]
  • 14.Underhill P.A., Poznik G.D., Rootsi S., Järve M., Lin A.A., Wang J., Passarelli B., Kanbar J., Myres N.M., King R.J., et al. The Phylogenetic and Geographic Structure of Y-Chromosome Haplogroup R1a. Eur. J. Hum. Genet. 2015;23:124–131. doi: 10.1038/ejhg.2014.50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Peričić M., Lauc L.B., Klarić I.M., Rootsi S., Janićijević B., Rudan I., Terzić R., Čolak I., Kvesić A., Popović D., et al. High-Resolution Phylogenetic Analysis of Southeastern Europe Traces Major Episodes of Paternal Gene Flow Among Slavic Populations. Mol. Biol. Evol. 2005;22:1964–1975. doi: 10.1093/molbev/msi185. [DOI] [PubMed] [Google Scholar]
  • 16.Semino O., Magri C., Benuzzi G., Lin A.A., Al-Zahery N., Battaglia V., Maccioni L., Triantaphyllidis C., Shen P., Oefner P.J., et al. Origin, Diffusion, and Differentiation of Y-Chromosome Haplogroups E and J: Inferences on the Neolithization of Europe and Later Migratory Events in the Mediterranean Area. Am. J. Hum. Genet. 2004;74:1023–1034. doi: 10.1086/386295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Battaglia V., Fornarino S., Al-Zahery N., Olivieri A., Pala M., Myres N.M., King R.J., Rootsi S., Marjanovic D., Primorac D., et al. Y-Chromosomal Evidence of the Cultural Diffusion of Agriculture in Southeast Europe. Eur. J. Hum. Genet. 2009;17:820–830. doi: 10.1038/ejhg.2008.249. Erratum in Eur. J. Hum. Genet. 2009, 17, 853. https://doi.org/10.1038/ejhg.2009.21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Olalde I., Brace S., Allentoft M.E., Armit I., Kristiansen K., Booth T., Rohland N., Mallick S., Szécsényi-Nagy A., Mittnik A., et al. The Beaker Phenomenon and the Genomic Transformation of Northwest Europe. Nature. 2018;555:190–196. doi: 10.1038/nature25738. Erratum in Nature 2018, 555, 543. https://doi.org/10.1038/nature26164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Rootsi S., Myres N.M., Lin A.A., Järve M., King R.J., Kutuev I., Cabrera V.M., Khusnutdinova E.K., Varendi K., Sahakyan H., et al. Distinguishing the Co-Ancestries of Haplogroup G Y-Chromosomes in the Populations of Europe and the Caucasus. Eur. J. Hum. Genet. 2012;20:1275–1282. doi: 10.1038/ejhg.2012.86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Rootsi S., Kivisild T., Benuzzi G., Help H., Bermisheva M., Kutuev I., Barać L., Peričić M., Balanovsky O., Pshenichnov A., et al. Phylogeography of Y-Chromosome Haplogroup I Reveals Distinct Domains of Prehistoric Gene Flow in Europe. Am. J. Hum. Genet. 2004;75:128–137. doi: 10.1086/422196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Rootsi S., Zhivotovsky L.A., Baldovič M., Kayser M., Kutuev I.A., Khusainova R., Bermisheva M.A., Gubina M., Fedorova S.A., Ilumäe A.-M., et al. A Counter-Clockwise Northern Route of the Y-Chromosome Haplogroup N from Southeast Asia towards Europe. Eur. J. Hum. Genet. 2007;15:204–211. doi: 10.1038/sj.ejhg.5201748. [DOI] [PubMed] [Google Scholar]
  • 22.Slatkin M. A Measure of Population Subdivision Based on Microsatellite Allele Frequencies. Genetics. 1995;139:457–462. doi: 10.1093/genetics/139.1.457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Ballantyne K.N., Goedbloed M., Fang R., Schaap O., Lao O., Wollstein A., Choi Y., van Duijn K., Vermeulen M., Brauer S., et al. Mutability of Y-Chromosomal Microsatellites: Rates, Characteristics, Molecular Bases, and Forensic Implications. Am. J. Hum. Genet. 2010;87:341–353. doi: 10.1016/j.ajhg.2010.08.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Popoiu G., Stanciu F., Cuțăr V., Vladu S., Podgoreanu P., Nicola V., Stoian I.M., Procopciuc A., Hațegan B., Negoiță B., et al. Population Genetics Data of 21 Autosomal STR Loci in the Romanian Population. Data. 2025;10:80. doi: 10.3390/data10060080. [DOI] [Google Scholar]
  • 25.Gouy A., Zieger M. STRAF—A Convenient Online Tool for STR Data Evaluation in Forensic Genetics. Forensic Sci. Int. Genet. 2017;30:148–151. doi: 10.1016/j.fsigen.2017.07.007. [DOI] [PubMed] [Google Scholar]
  • 26.Athey T.W. Haplogroup Prediction from Y-STR Values Using a Bayesian-Allele- Frequency Approach. J. Genet. Geneal. 2006;2:34–39. [Google Scholar]
  • 27.Nei M. Molecular Evolutionary Genetics. Columbia University Press; New York, NY, USA: 1987. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ijms-27-07121-s001.zip (451.9KB, zip)

Data Availability Statement

The original contributions presented in this study are included in the article and Supplementary Materials. Further inquiries can be directed to the corresponding authors.

Articles from International Journal of Molecular Sciences are provided here courtesy of Multidisciplinary Digital Publishing Institute (MDPI)

RESOURCES