Population Genetics Data of 21 Autosomal STR Loci in the Romanian Population

Abstract

This study aimed to determine the allele frequencies and genetic diversity of 21 autosomal short tandem repeat (STR) loci from the Expanded U.S. Core Loci and European Standard Set in the Romanian population. A random sample of 928 unrelated men from all Romanian counties was analyzed using the Investigator 24plex QS and Investigator 24plex GO! Kits (Qiagen). The genotypes were determined, and the allele frequencies were calculated using the STRidER tool. The results provide updated population genetic data for the Romanian population, which is essential for accurate calculation of DNA evidence weight in forensic casework.

1. Introduction

In recent decades, the field of forensic sciences has undergone a series of changes, by introducing new specializations and improved methods, and also by removing those that did not have a solid scientific basis (e.g., bite mark analysis [1], microscopic hair analysis [2]).

Forensic genetics was one of the specializations that revolutionized the field of forensic sciences, which in the last 30 years has experienced enormous development and ”has produced valuable evidence that has contributed to the successful prosecution and conviction of criminals as well as to the exoneration of innocent people” [3].

Nowadays, the vast majority of forensic genetics laboratories use data obtained as a result of analyzing a set of short tandem repeat (STR) markers that together form a genetic profile. Depending on the number of markers contained and the rarity of the alleles obtained, the genetic profile may be unique (except for monozygotic twins where other techniques need to be applied [4]).

In the context of the freedom of movement, unprecedented in all human history, and the criminal challenges it poses (e.g., terrorism and cross-border crime [5,6]), law enforcement agencies have established the use of standard sets of STR markers, like Expanded U.S. Core Loci or European Standard Set and the interconnection of national forensic DNA databases.

When a genetic profile obtained from a person (also called reference/known sample) is identical to the genetic profile obtained from an object related to a crime (also called unknown sample/casework sample), we say there is a match. These matches can be the result of direct comparisons of two DNA profiles obtained from a casework sample and a reference sample, or the result of comparisons between a DNA profile obtained from a casework sample and one or more forensic DNA databases. Matches of two or more persons are also relevant to forensic work when it comes to investigative steps, kinship, pedigree, or familial DNA searches. Likewise, matches of two or more DNA profiles obtained from casework can be used for linking different cases.

In forensic genetics, as a general rule, when there is a match between a reference sample and an unknown sample, the weight of evidence must be determined [7]. To calculate it, data on the relative frequency of alleles in the population are required [8]. Although a number of population genetic studies have been published over the years [9,10,11,12,13,14,15,16,17,18], given the abrupt decline in Romania’s population (every 10 years in the last three decades the population has decreased by approximately 1 million) [18], the immigration and the increased number of autosomal STR markers that are currently used in forensic genetics, we considered it appropriate to carry out a new population genetic study, which would reflect the current Romanian population status.

In this study we optimized our data analysis protocols and utilized the place of birth as the benchmark for selecting the population group.

Another particularity of this study is that only samples from men, who do not share the same genetic haplotype of the Y chromosome, were used, using an additional step to ensure that there is no kinship between the participants.

2. Materials and Methods

For this study we randomly selected a sample of n = 928 men from all Romanian counties respecting, for most of the counties, the corresponding proportion of Romania’s population, according to 2021 population census [19], as follows in

Table 1. Distribution of analyzed samples.

County	Percent of the Romanian Population (According to 2021 Population Census)	Number of Analyzed Biological Samples
Alba	1.71%	7 *
Arad	2.15%	22
Argeș	2.99%	30
Bacău	3.16%	32
Bihor	2.89%	29
Bistrița-Nasăud	1.55%	16
Botoșani	2.06%	21
Brașov	2.87%	29
Brăila	1.48%	15
Buzău	2.13%	21
Caraș-Severin	1.29%	13
Călărași	1.49%	5 *
Cluj	3.56%	36
Constanța	3.44%	34
Covasna	1.05%	10
Dâmbovița	2.52%	25
Dolj	3.15%	30 *
Galați	2.61%	26
Giurgiu	1.38%	8 *
Gorj	1.65%	17
Harghita	1.53%	13 *
Hunedoara	1.90%	19
Ialomița	1.32%	13
Iași	3.99%	40
Ilfov	2.85%	4 *
Maramureș	2.37%	24
Mehedinți	1.23%	12
Mureș	2.72%	27
Neamț	2.38%	24
Olt	2.01%	20
Prahova	3.65%	36
Satu Mare	1.74%	4 *
Sălaj	1.11%	11
Sibiu	2.04%	20
Suceava	3.37%	34
Teleorman	1.70%	17
Timiș	3.41%	28 *
Tulcea	1.01%	10
Vaslui	1.97%	20
Vâlcea	1.79%	18
Vrancea	1.76%	18
Bucharest City	9.01%	90

* Counties where target population size was not reached.

.

The biological samples, consisting of epithelial buccal cells, were collected from convicted persons through non-invasive methods by medical personnel, as a result of applying the national law no. 76 from 2008 [20]. EasiCollect^® Plus devices (Qiagen, Venlo, The Netherlands) were used to collect 96.65% of the samples, while 3.35% were collected using different types of buccal swabs.

After collection, the samples were processed in one of the following ways, depending on the type of support or how well the sample was preserved:

(a)

DNA was extracted with QIAamp Investigator BioRobot Kit (Qiagen) on BioRobot Universal System (Qiagen) using the kit manufacturer’s protocol. The DNA extraction was followed by STR amplification with the Investigator 24plex QS kit (Qiagen) and a Veriti™ Thermal Cycler (ThermoFisher Scientific, Waltham, MA, USA) for 21 autosomal STR (A-STR) loci (

Table 2. A-STR loci analyzed and their chromosomal locations [21].

Locus	Chromosomal Location
CSF1PO	5q32
D1S1656	1q42.2
D2S441	2p14
D2S1338	2q35
D3S1358	3p21.31
D5S818	5q23.2
D7S820	7q21.11
D8S1179	8q24.13
D10S1248	10q26.3
D12S391	12p13.2
D13S317	13q31.1
D16S539	16q24.1
D18S51	18q21.33
D19S433	19q12
D21S11	21q21.1
D22S1045	22q12.3
FGA	4q31.3
SE33	6q15
TH01	11p15.5
TPOX	2p25.3
vWA	12p13.31

), one Y chromosome STR (Y-STR) locus (DYS391), Amelogenin, and two quality controls, using the kit manufacturer’s protocol.

(b)

DNA was extracted with QIAsymphony DNA Investigator Kit (Qiagen) and QIAsymphony SP/AS instrument (Qiagen) using the kit manufacturer’s protocol. The DNA was quantified with Investigator Quantiplex Pro RGQ (Qiagen) and RotorGene Q Real-Time System (Qiagen) using the kit manufacturer’s protocol. PCR amplification was made with the Investigator 24plex QS kit (Qiagen) and a Veriti™ Thermal Cycler (ThermoFisher Scientific) for 21 A-STR loci (Table 2), one Y-STR locus (DYS391), Amelogenin, and two quality controls, using the kit manufacturer’s protocol.

(c)

DNA was extracted with Investigator STR GO! Lysis Buffer (Qiagen) using the kit manufacturer’s protocol. The DNA extraction was followed by STR amplification with the Investigator 24plex GO! Kit (Qiagen) and a Veriti™ Thermal Cycler (ThermoFisher Scientific) for 21 A-STR loci (Table 2), one Y-STR locus (DYS391), Amelogenin, and two quality controls, using the kit manufacturer’s protocol.

(d)

For some samples, no separate DNA extraction step was performed. Instead, the samples were subjected directly to PCR amplification using the Investigator 24plex GO! Kit (Qiagen) and a Veriti™ Thermal Cycler (ThermoFisher Scientific). This amplification targeted 21 A-STR loci (Table 2), one Y-STR locus (DYS391), the Amelogenin locus, and two quality control markers. The amplification was carried out following the manufacturer’s protocol.

Regardless of how the samples were processed in the previous steps, the detection and separation of PCR products were carried out using ABI 3500 Genetic Analyzer (ThermoFisher Scientific) according to the recommendations of the amplification kit manufacturer.

Genotypes were determined using GeneMapper ID-X 1.6 with the default settings from Analysis_HID_3500_200rfu protocol.

Statistical analysis was performed with quality control of autosomal Short Tandem Repeat allele frequency databasing (STRidER) v3/R2 [22].

The calculations for observed heterozygosity, expected heterozygosity, Hardy–Weinberg equilibrium exact test, matching probability, power of discrimination, polymorphism information content, power of exclusion, and typical paternity index were performed using STRAF 2.2.2 [23].

3. Results

The allele’s frequency for the 21 STR loci from the Expanded U.S. Core Loci and European Standard Set studied in the Romanian population and statistical parameters are summarized in

Table 3. Alleles frequency, observed heterozygosity (Ho), expected heterozygosity (He), Hardy–Weinberg equilibrium exact test (P), matching probability (PM), power of discrimination (PD), polymorphism information content (PIC), power of exclusion (PE), and typical paternity index (TPI) for 11 STR markers.

Allele	CSF1PO	D1S1656	D2S441	D2S1338	D3S1358	D5S818	D7S820	D8S1179	D10S1248	D12S391	D13S317
7						0.003	0.023				0.001
8	0.003					0.001	0.151	0.023			0.141
9	0.030		0.001			0.032	0.148	0.012			0.094
10	0.269	0.001	0.170			0.079	0.249	0.074	0.001		0.061
11	0.315	0.106	0.321		0.001	0.310	0.244	0.072	0.002		0.309
11.3			0.086
12	0.316	0.126	0.037		0.001	0.378	0.156	0.121	0.021		0.270
12.3			0.001
13	0.057	0.102	0.033		0.002	0.184	0.025	0.307	0.228		0.098
13.3			0.001
14	0.008	0.074	0.303		0.092	0.010	0.004	0.237	0.287	0.001	0.026
14.3		0.001
15	0.002	0.161	0.040		0.288	0.003		0.118	0.249	0.030	0.001
15.3		0.045
16		0.147	0.008	0.043	0.275			0.032	0.164	0.023
16.3		0.036
17		0.038		0.195	0.192			0.003	0.041	0.095
17.3		0.108								0.008
18		0.004		0.093	0.138			0.001	0.006	0.203
18.3		0.039								0.022
19				0.102	0.010					0.131
19.3		0.010								0.011
20				0.159	0.001					0.125
20.2		0.001
20.3		0.002								0.001
21				0.030						0.106
22				0.024						0.116
23				0.135						0.082
24				0.105						0.030
25				0.099						0.014
26				0.012						0.004
27				0.002
28				0.001
HO	0.719	0.886	0.756	0.869	0.795	0.710	0.798	0.819	0.778	0.880	0.777
HE	0.724	0.891	0.765	0.875	0.776	0.719	0.808	0.808	0.774	0.883	0.789
P	0.094	0.684	0.658	0.788	0.157	0.610	0.382	0.788	0.629	0.470	0.417
PM	0.128	0.022	0.090	0.029	0.088	0.125	0.064	0.065	0.088	0.026	0.075
PD	0.872	0.978	0.910	0.971	0.912	0.875	0.936	0.935	0.912	0.974	0.925
PIC	0.673	0.881	0.730	0.862	0.741	0.671	0.780	0.784	0.738	0.871	0.759
PE	0.458	0.766	0.521	0.732	0.590	0.444	0.596	0.635	0.559	0.756	0.557
TPI	1.778	4.377	2.053	3.803	2.442	1.725	2.481	2.762	2.252	4.180	2.242

and

Table 4. Alleles frequency, observed heterozygosity (Ho), expected heterozygosity (He), Hardy– Weinberg equilibrium exact test (P), matching probability (PM), power of discrimination (PD), polymorphism information content (PIC), power of exclusion (PE), and typical paternity index (TPI) for 10 STR markers.

Allele	D16S539	D18S51	D19S433	D21S11	D22S1045	FGA	SE33	TH01	TPOX	vWA
6								0.263
7								0.138	0.001
8	0.019							0.125	0.526
8.3								0.001
9	0.115	0.001						0.199	0.103
9.3								0.265
10	0.074	0.006			0.001		0.001	0.009	0.063
11	0.307	0.016	0.002		0.134		0.001	0.001	0.281
12	0.292	0.110	0.095		0.013		0.006		0.026
12.2			0.005
13	0.165	0.138	0.246		0.003		0.005			0.001
13.2			0.013				0.005
14	0.026	0.175	0.324		0.063		0.025			0.124
14.2			0.033				0.002
15	0.002	0.149	0.152		0.381		0.030			0.100
15.2		0.001	0.041
15.3							0.001
16		0.152	0.047		0.311	0.001	0.040			0.193
16.2			0.032				0.001
16.3							0.001
17		0.120	0.002		0.079	0.001	0.077			0.296
17.1							0.001
17.2			0.006
17.3							0.001
18		0.054			0.008	0.017	0.083			0.193
18.2			0.001
18.3							0.001
19		0.039			0.006	0.079	0.095			0.077
19.2			0.001				0.002
19.3						0.001
20		0.022				0.118	0.072			0.016
20.2						0.001	0.006
21		0.010				0.171	0.028
21.2						0.003	0.009
22		0.005				0.171	0.007
22.2						0.015	0.031
23		0.002				0.150	0.002
23.2						0.009	0.034
24		0.001				0.148	0.001
24.2						0.002	0.030
25				0.001		0.080
25.2						0.002	0.030
26				0.003		0.025
26.2							0.037
27				0.018		0.006	0.001
27.2							0.055
28				0.158		0.001
28.1				0.001
28.2							0.096
28.3							0.001
29				0.237			0.007
29.2				0.004			0.070
30				0.185			0.001
30.2				0.048			0.036
31				0.060
31.2				0.113			0.032
32				0.008			0.001
32.2				0.122			0.023
33							0.002
33.1				0.001
33.2				0.039			0.008
34							0.003
34.2				0.004			0.002
35							0.002
35.2				0.001
36							0.001
38							0.001
HO	0.754	0.864	0.792	0.832	0.722	0.893	0.947	0.776	0.643	0.801
HE	0.773	0.873	0.796	0.850	0.729	0.869	0.943	0.786	0.629	0.806
P	0.453	0.226	0.654	0.125	0.766	0.987	0.002	0.981	0.483	0.121
PM	0.085	0.030	0.070	0.041	0.116	0.032	0.007	0.079	0.190	0.064
PD	0.915	0.970	0.930	0.959	0.884	0.968	0.993	0.921	0.810	0.936
PIC	0.739	0.860	0.770	0.832	0.687	0.855	0.941	0.752	0.574	0.780
PE	0.517	0.723	0.584	0.660	0.463	0.782	0.892	0.555	0.346	0.600
TPI	2.035	3.683	2.404	2.974	1.798	4.687	9.469	2.231	1.402	2.508

(STRidER reference code STR000441).

The numbers of alleles observed for every locus analyzed is presented in

Table 5. Numbers of alleles observed for every locus analyzed.

Locus	Number of Alleles
CSF1PO	8
D1S1656	17
D2S441	11
D2S1338	13
D3S1358	10
D5S818	9
D7S820	8
D8S1179	11
D10S1248	9
D12S391	17
D13S317	9
D16S539	8
D18S51	17
D19S433	15
D21S11	17
D22S1045	10
FGA	20
SE33	49
TH01	8
TPOX	6
vWA	8

and varied from six alleles for TPOX marker to 49 for SE33 marker, with an overall mean value of 13 alleles per locus.

No significant differences were seen between observed heterozygosity and expected heterozygosity.

With the exception of the SE33 locus (P = 0.002), no deviations from Hardy–Weinberg equilibrium were observed. This may be due to the high mutation rate of this locus, the highest of all the markers analyzed (0.64%), more than twice as high as the second highest mutation rate [24].

The lowest random match probability (PM) value is at locus SE33 (PM = 0.007) and the highest at TPOX (PM = 0.190) as shown in Figure 1, which means that TPOX marker has the lowest power of discrimination of all 21 analyzed markers (PD = 0.810), while the SE33 marker has the highest power of discrimination (PD = 0.993).

In the case of genetic relationship, the power of exclusion (PE) and typical paternity index (TPI) show the same distribution of the markers, with TPOX locus having the lowest value and the SE33 locus with the highest one (Figure 2).

4. Discussion

Comparing the number of alleles observed in the Romanian population in this study with the number of alleles discovered worldwide [21], it results that a proportion varying between 17% and 48% of these alleles are found in the Romanian population (Figure 3), with loci D22S1045 (48%), D10S1248 (47%), and SE33 (45%) having the highest percent of the alleles found worldwide and the loci FGA (17%) and D7S820 (22%) having the lowest percent of the alleles found worldwide.

The comparative analysis of similar studies in the Romanian population highlighted slight differences, when it comes to obtained alleles and their frequencies. However, the differences between the obtained results for this sample and those obtained by Stanciu et al. [16] were minor, with the maximum calculated differences between allele frequencies per locus being in the range of 1.2–4.2%. Alleles obtained in this study and not in the other one, and vice versa, are considered very rare alleles because they were observed less than five times in the samples analyzed [25]. The main difference between the two analyses lies in the number of markers that were processed, six more in this study than in the previous one.

More significant differences were observed when comparing the results of this study and those obtained by Barbarii et al. [11]. Thus, six alleles with frequency differences of more than 5% were highlighted. Also, a more diverse distribution of alleles was shown in our sample, with a difference of more than 40 alleles observed in the new study. These differences could be explained by the large difference in the size of the sample (928 vs. 243), the way the sample was selected (all Romania’s counties vs Bucharest area) and the 20-year difference between the two studies.

5. Conclusions

In this study, following the latest international guidelines when it comes to the field of forensic genetics, we obtained a set of allele frequencies suitable for the purpose of making biostatistical calculations which can be used for forensic DNA identification, paternity testing, and probabilistic interpretation to evaluate STR DNA profiles from a mixture of contributors.

Moreover, the results obtained by comparing the data presented here with those obtained in similar studies, of the population of other countries, can contribute to a better understanding of some aspects regarding population migration throughout history across Europe and other continents.