Bootstrapping estimation for estimating relative potency in combinations of bioassays

Estimation of relative potency in biological assay is very important in pharmaceutical and toxicological quality control. The statistical procedure to estimate this parameter produces a ratio of two other statistics resulting complexity for its distribution when inference will be made. In this paper, a bootstrapping procedure is proposed to approximate its population distribution and hence the estimation for this parameter in the combinations of bioassays. The proposed method overcomes the problems with distribution assumptions. We conclude with a comparison of this procedure to the well-known weighted mean method for a real data set.