Estimation of heterozygosity for single-probe multilocus DNA fingerprints |
| |
Authors: | Stephens, JC Gilbert, DA Yuhki, N O'Brien, SJ |
| |
Affiliation: | Laboratory of Viral Carcinogenesis, Frederick Cancer Research and Development Center, Maryland 21702-1201. |
| |
Abstract: | In spite of the increasing application of DNA fingerprinting to naturalpopulations and to the genetic identification of humans, explicit methodsfor estimation of basic population genetic parameters from DNAfingerprinting data have not been developed. Contributing to this omissionis the inability to determine, for multilocus fingerprinting probes,relatively important genetic information, such as the number of loci, thenumber of alleles, and the distribution of these alleles into specificloci. One of the most useful genetic parameters that could be derived fromsuch data would be the average heterozygosity, which has traditionally beenemployed to measure the level of genetic variation within populations andto compare genetic variation among different loci. We derive here explicitformulas for both the estimation of average heterozygosity at multiplehypervariable loci and a maximum value for this estimate. These estimatesare based upon the DNA restriction-pattern matrices that are typical forfingerprinting studies of humans and natural populations. For severalempirical data sets from our laboratory, estimates of average and maximalheterozygosity are shown to be relatively close to each other. Furthermore,variances of these statistics based on simulation studies are relativelysmall. These observations, as well as consideration of the effect ofmissing alleles and alternate numbers of loci, suggest that the averageheterozygosity can be accurately estimated using phenotypic DNA fingerprintpatterns, because this parameter is relatively insensitive to the lack ofcertain genetic information. |
| |
Keywords: | |
本文献已被 Oxford 等数据库收录! |
|