Genetic parameters of five new European Standard Set STR loci (D10S1248, D22S1045, D2S441, D1S1656, D12S391) in the population of eastern Croatia

Aim To establish allele frequencies and genetic parameters in eastern Croatia population and to compare them with those in other populations. The second aim was to compare the genetic profiles obtained with different forensic kits amplifying the same genetic markers. Methods Blood samples of 217 unrelated individuals from eastern Croatia were genotyped using AmpFlSTR NGM kit. Allele distribution and other genetic parameters were determined for 15 short tandem repeat (STR) loci, including the 5 loci recently added to the European Standard Set (ESS) of STR loci (D10S1248, D22S1045, D2S441, D1S1656, and D12S391). Ninety-six samples underwent duplicate analysis using AmpFlSTR Identifiler kit. Results Power of discrimination was highest for the two new ESS loci, D1S1656 (0.97254) and D12S391 (0.97339). Comparison of allele frequencies for 5 new ESS loci in our sample with previously published population data showed a significant difference from Maghreb population on D2S441 and from American Caucasian population on D1S1656. Comparison of allele frequencies for standard 10 STR loci with all the neighboring populations’ data showed a significant difference only from Albanian population (on D2S1338, D18S51, and TH01). Discordant genotypes were observed in 5 (5.2%) samples at a single locus when amplified with both AmpFlSTR NGM and AmpFlSTR Identifiler kit. Conclusion New ESS STR loci are highly polymorphic and short, and therefore very useful for the analysis of challenging forensic samples. DNA samples purposed for establishing databases should be routinely amplified in duplicate.

Aim To establish allele frequencies and genetic parameters in eastern Croatia population and to compare them with those in other populations. The second aim was to compare the genetic profiles obtained with different forensic kits amplifying the same genetic markers.
Methods Blood samples of 217 unrelated individuals from eastern Croatia were genotyped using AmpFlSTR NGM kit. Allele distribution and other genetic parameters were determined for 15 short tandem repeat (STR) loci, including the 5 loci recently added to the European Standard Set (ESS) of STR loci (D10S1248, D22S1045, D2S441, D1S1656, and D12S391). Ninety-six samples underwent duplicate analysis using AmpFlSTR Identifiler kit.

Results
Power of discrimination was highest for the two new ESS loci, D1S1656 (0.97254) and D12S391 (0.97339). Comparison of allele frequencies for 5 new ESS loci in our sample with previously published population data showed a significant difference from Maghreb population on D2S441 and from American Caucasian population on D1S1656. Comparison of allele frequencies for standard 10 STR loci with all the neighboring populations' data showed a significant difference only from Albanian population (on D2S1338, D18S51, and TH01). Discordant genotypes were observed in 5 (5.2%) samples at a single locus when amplified with both AmpFlSTR NGM and AmpFlSTR Identifiler kit.
Conclusion New ESS STR loci are highly polymorphic and short, and therefore very useful for the analysis of challenging forensic samples. DNA samples purposed for establishing databases should be routinely amplified in duplicate.
To facilitate DNA profiles comparison between databases of different European countries, The European Network of Forensic Science Institutes (ENFSI) and European DNA Profiling Group (EDNAP) have recently added five new loci (D10S1248, D22S1045, D2S441, D1S1656, and D12S391) to the European Standard Set of short tandem repeat (STR) loci (1,2). These new loci were included into the AmpFlSTR NGM PCR amplification kit (NGM kit; Life Technologies, Foster City, CA, USA).
The Laboratory for DNA Analysis in Osijek was established to participate in the identification of missing persons after the war in Croatia (1991)(1992)(1993)(1994)(1995). In collaboration with the laboratories in Zagreb and Split, a database of genotypes of missing persons' relatives was created including approximately 5000 persons. The greatest part of the included genetic information is based on the 15 loci incorporated in AmpFlSTR Identifiler PCR amplification kit (Identifiler kit; Life Technologies, Foster City, CA, USA). Skeletal remains are identified by comparing the genotype of each piece of skeletal remains with the genotypes in the missing persons' relatives database. Such non-targeted matching in a database containing several thousands genotypes considerably decreases the reliability of the established match. Still, the majority of identified skeletal remains were matched in such a way, as genotypes of the missing persons from the father-mother-child trio. Even within so large a database, hundreds of genotypes of skeletal remains still do not have a match, due to a lack of adequate relatives. Matching a profile created from a piece of skeletal remains across the whole database returns many adventitious matches, partly because some genotyped loci have low discrimination power (2). An especially large number of adventitious matches is present if the genetic profile from skeletal remains is partial.
A targeted approach to DNA typing, loci on the Y-chromosome and mtDNA can be amplified, but at a database level more useful are the loci on somatic chromosomes. Evidential value of a genetic match based on STR typing relies on high polymorphism and a large number of STR loci. In order to obtain as much as possible genetic information, we used the NGM kit. Our aim was to increase the number of genetic markers in order to achieve higher evidential value of STR typing and to amplify short STR loci, often better preserved in degraded samples. Especially valuable are three new "mini" STR loci (D10S1248, D22S1045, and D2S441), engineered to produce short amplicons (up to 150 bp) that are more successfully obtained from the most degraded samples. The remaining two new loci (D1S1656 and D12S391) are also relatively short and highly polymorphic (3-7). Besides obtaining information on the 5 new loci, the NGM kit in-cludes improved chemistry that maximizes performance on challenging samples.
In the new European Standard Set (ESS) of STR loci, allele distribution and genetic parameters still have to be determined. A population study on the new loci has been performed for several countries (including Belgium, Germany, Hungary, Maghreb countries, Poland, and USA) (7-12). However, there has been no such study either for Croatian or its neighboring populations. Therefore, we carried out a population study on a sample from eastern Croatia, which might be the most appropriate regional sample, because this part of the country sustained the greatest human losses during the war. Since the greatest part of our relatives' database is based on the Identifiler kit, which shares 10 loci with NGM kit (D3S1358, vWA, D16S539, D2S1338, D8S1179, D21S11, D18S51, D19S433, TH01, and FGA), we compared the genetic profiles obtained with both of these kits amplifying the same genetic markers. We also compared the obtained genetic parameters for 15 STR loci with the available population data from the neighboring countries.

Sampling and extraction
This study was conducted on a population sample of 217 unrelated persons from eastern Croatia (population of approximately 500 000 inhabitants). A study based on 100-150 participants is generally accepted as appropriate for determining population data (13).    locus (0.03974) was observed, but rejected after Boniferroni correction (Table 1). For five new ESS loci, our sample showed a significant difference in allele frequencies only from Maghreb population on D2S441 (P < 0.001) and from American Caucasian population on D1S1656 (P = 0.003) ( Table 2). For the remaining 10 loci, allele frequencies included in NGM kit were compared with the previously published population data for Croatia, Albania, Bosnia, Hungary, Macedonia, Serbia and Montenegro, and Slovenia and significant differences were found for Albanian population on the locus D2S1338 (P = 0.022), D18S51 (P = 0.029), and TH01 (P = 0.006) ( Table 3).

DiSCuSSion
We found that new ESS STR loci in Croatian population were highly polymorphic, which is in line with the previously published data about these loci (7-12).
When allele frequencies of 10 standard ESS loci for Croatian population were compared with Albania, Bosnia, Hungary, Macedonia, Serbia and Montenegro, and Slovenia (15-21), significant differences were found only for Albanian population. Such lack of significant differences is in line with the history of population migrations on the Balkans in the last centuries (22,23).
Since there were no significant differences in the allele frequencies of 10 STR loci between our and the neighboring populations, it might be concluded that there are no significant differences for the new STR loci too. This assumption should be tested in future research, but due to lack of population data for the new ESS loci, our population data might serve as a rough approximation of the population data in these countries.
Analysis of more STR loci increases the discrimination power. Independent assortment of STR markers on the same chromosome cannot be assumed and a complex kinship analysis requires an increased number of loci across different chromosomes. Therefore, in case of syntenic loci it is recommended to use a more informative locus in a specific case of probability calculation (12). The NGM kit includes two pairs of syntenic loci, vWA and D12S391 on the chromosome 12, and D2S1338 and D2S441 on the chromosome 2. When syntenic loci were excluded from the kinship analysis, the probability calculation was based on 13 out of 15 loci. The probability calculation based on 13 loci in non-targeted matching approach might be insufficient for reliable identification, ie, in cases of reverse uniparental test- ing, especially if mutation event is also included in the calculation. When samples were amplified with both NGM and Identifiler kit, 16 non-syntenic out of 20 loci were used for probability calculation. DNA profiling based on 20 loci has very high discriminatory power (24) and should enable reliable identification even in the complex relationship testing.
Our relatives' database is based on 15 loci included in the Identifiler kit, 10 of which are also included in the NGM kit. In order to identify discordant genotypes, we performed a duplicate analysis for 96 samples. We observed 5 (5.2%) discordant genotypes and in each case allelic dropout occurred when samples were amplified with NGM kit. Nonamplification of the second allele in a heterozygous genotype was observed for the D18S51 (once), D2S1338 (twice), and D21S11 (twice), and it was probably a result of random nature of PCR. Therefore, DNA samples purposed for establishing genotype databases should be routinely amplified in duplicate.
In a non-targeted approach to matching, when the genotype of each piece of skeletal remains is compared to thousands of genotypes in the database, false inclusion is common. In our experience, in about 5% of the cases more than 5 potential parents or children match the genotype of a specific piece of skeletal remains on 15 loci. Experience taught us not to "believe" the DNA without correspondence with other forensic information, like place of disappearance or matching personal traits or belongings. Nontargeted matching through large databases should never rely solely on genetic information, and anthropologic or other forensic data should be taken into account.
acknowledgment We thank Iris Banjan, Sunčica Jurušić, and Ksenija Andrić for language editing of the manuscript. Declaration of authorship GĆ was in charge of technical organizational aspects, performed a part of the experimental work, interpreted the data, and worked on the manuscript from the early beginning until the final version. VG performed a part of the experimental work, conducted statistical analysis, interpreted the data, and participated in writing of the manuscript. VP fully participated in composing and writing of the manuscript and performed a part of the experimental work. DS performed a part of the experimental work, interpreted the data, and participated in manuscript design, preparation, and review.