[Tests for Hardy-Weinberg equilibrium in Japanese population]

In population genetics, the absence of the departure from Hardy-Weinberg equilibrium (HWE) is usually tested when a population study of a certain DNA marker is performed to show the observed allele frequencies represent those of the whole population. The goodness-of-fit test (chi 2 test) assuming chi 2 approximation has frequently been used with classical blood type markers having a few alleles. However, new tests suitable for DNA markers having many alleles, such as homozygosity test, likelihood ratio test and Guo-Thompson's (G-T') exact test, have recently been devised. In the present study, appropriate tests for HWE was studied using population data of 206 Japanese individuals with 9 different short tandem repeat loci. Firstly, we found that the recommendation of NRC II for the treatment of rare allele frequencies (If a bin in the database contains fewer than five entries, it is pooled with adjacent bins so that no bin has fewer than five) is quite reasonable for personal identification in forensic sciences. Secondly, we proposed that homozygosity test, likelihood ratio test and G-T's exact test should be applied altogether and HWE of the sample population should be valid only when all of the three tests were cleared.