Sequential Bayesian approach for genetic diversity analysis of the piracanjuba fish (Brycon orbignyanus)

Abstract

In the sequential Bayesian approach, the sample size is not fixed before the experiment; it is determined based on the observations made. The procedure concludes when there is enough information to estimate the parameters, according to a stopping criterion. A parameter of interest in population genetics is the proportion obtained from the allele frequency at one or more loci to verify Hardy-Weinberg equilibrium (HWE). The objective of this study was to assess the occurrence of HWE in a population of piracanjuba fish (Brycon orbignyanus) by estimating the allele proportion and expected genotype proportions using a sequential Bayesian approach. Additionally, a comparison was made with frequentist and Bayesian approaches. Initially, genotypic profiles were analyzed at a microsatellite DNA locus, Bh6, in 49 fish, to determine the frequency of observed alleles and genotypes at the UFSJ Genetic Resources Laboratory. Seven allele classes were observed; thus, under the assumption of sampling independence, the likelihood is multinomial. The estimation of allele and genotype proportions was then carried out using frequentist, Bayesian, and sequential Bayesian approaches. A uniform prior and a cost of 10–3 were considered. The estimates from the three approaches were compared, and it was concluded that the sequential approach proved effective, utilizing only 55.1% of the available data, thereby reducing the sample size and optimizing the procedure. Using a chi-square test at a 5% probability level, it was concluded that the studied sample is in Hardy-Weinberg equilibrium (p-value: 0.9800245).