A new method for the evaluation of matches in non-recombining genomes: application to Y-chromosomal short tandem repeat (STR) haplotypes in European males

A 9-locus microsatellite framework (minimal haplotype), previously develope d for forensic purposes so as to facilitate stain analysis, personal identi fication and kinship testing, has been adopted for the establishment of a l arge reference database of male European Y-chromosomal haplotypes. The exte nt of population stratification pertaining to this database, an issue cruci al for its practical forensic application, was assessed through analysis of molecular variance (AMOVA) of the 20 regional samples included. Despite th e notion of some significant haplotype frequency differences, which were fo und to correlate with known demographic and historic features of Europeans, AMOVA generally revealed a high level of genetic homogeneity among the pop ulations analyzed. Owing to their high diversity, however, accurate frequen cy estimation is difficult for Y-STR haplotypes when realistic (i.e. modera tely sized) datasets are being used. As expected, strong pair-wise and high er order allelic associations were found to exist between all markers studi ed, implying that haplotype frequencies cannot be estimated as products of allele frequencies. A new extrapolation method was therefore developed whic h treats haplotype frequencies as random variables and generates estimates of the underlying distribution functions on the basis of closely related ha plotypes. This approach, termed frequency 'surveying', is based upon standa rd population genetics theory and can in principle be applied to any combin ation of markers located on the Y-chromosome or in the mitochondrial genome . Application of the method to the quality assured reference Y-STR haplotyp e database described herein will prove very useful for the evaluation of po sitive trace-donor matches in forensic casework. (C) 2000 Elsevier Science Ireland Ltd. All rights reserved.