Genetic analysis of a Sicilian population using 15 short tandem repeats

Human Biology, Apr 2003 by Calo, C M, Garofano, L, Mameli, A, Pizzamiglio, M, Vona, G

Results

Allele frequencies with markers under scrutiny in relation to the standard errors are shown in Table 2. The distribution of allele frequencies is unimodal for loci D8S1179, CSF1PO, and D5S818; multimodal for the rare alleles in loci D21S11 and FIBRA; and bimodal for all other alleles examined. Remarkably, within the Alia sample the presence of a rare allele (33.1) for locus D21S11 was discovered. In the samples there appear two heterozygote individuals (2%), which have not yet been observed in any other European population. The markers D2S1338 and D19S433, which have never before been used in research in population genetics and have only recently been put to use in forensic medicine (Garofano et al. 1999a; Garofano at al. 1999b), appear extremely polymorphic, showing 9 and 10 alleles each.

The exact test based on Markov chain analysis shows that all loci except TH01 are in Hardy-Weinberg equilibrium (Table 3). Of the 15 markers studied, two (TPOX and D2S1338) are located on chromosome 2 and two (CSF1PO and D5S818) are located on chromosome 15. The presence of linkage disequilibrium in each pair of loci located on the same chromosome was verified. The results, not shown, gave no significant values for either pair of loci.

The observed and expected heterozygosity as well as the PIC values for our samples are listed in Table 3. All loci had a high degree of heterozygosity, varying between 61.2% and 93.9% for TH01 and D18S51, respectively. Within the five markers D2S1338, D5S818, D7S820, THOl, and VWA, a decrease of heterozygosity had been observed and as a consequence an increase of homozygosity was observed. The polymorphic information content (PIC) was also calculated for each locus, except for loci TPOX and CSF1PO. All STRs showed PIC values ^gt;0.7, which is the limit value for use in linkage analysis (Hearne at al. 1992). Through the Bottleneck program (Cornuet and Luikart 1996), we discovered the indication of a recent bottleneck in the Alia population. Under the assumption of the Infinite Allele Model (IAM), all tests performed through the Bottleneck program (the sign, standardized differences, and Wilcoxon tests) showed a significant excessive heterozygosity in the Alia samples (p = 0.00001 and 0.00003, respectively). The data from Alia have been compared with those from seven other Sicilian regions, Agrigento, Caltanissetta, Catania, Messina, Palermo, Ragusa, and Siracusa (Barbara et al. 2000). Because of insufficient resources it was possible to compare only nine loci (D3S1358, D5S818, FIBRA, D7S820, D13S317, D18S51, D21S11), for a total of 93 alleles. G^sub ST^ values were then calculated in order to test for genetic diversity within Sicily (Table 3).

The genetic differentiation coefficient according to Nei (1973), G^sub ST^, showed the highest value for locus D21S11, while the lowest values were for VWA and D8S1179.

Genetic distances between Alia and other Sicilian populations were obtained by use of the Nei method (1972) (Table 4). They vary from a minimum of 0.0263 (Agrigento-Caltanissetta) to a maximum of 0.1993 (Catania-Palermo). Alia, in particular, showed the smallest distance from Ragusa and the largest from Palermo, despite the fact that it lies in the Palermo region. From the distances gathered from the matrix, a genetic tree was constructed using the neighbor-joining method (Saitou and Nei 1978) (Figure 1). The tree shows a clustered formation produced by Messina, Siracusa, and Catania, all located on the eastern coast of the island. The other four populations are located in two other distinct bunches, while Alia is not associated with any of the populations compared. From this previous comparison, other analyses were carried out using the data from the nine markers to obtain a total of 98 alleles, some of which have not been found in the Sicilian samples.