Original Articles

Parabola bioassays in the comparison of drug effects

Duan-zheng Xu


Canonical analysis procedures for the estimation of relative potencies when the log dose-response relationship is linear are available in standard texts, but cases where curved relationships are obtained can arise in vitamin B12 and gonadotrophin assays. The response curve of these assays can often be well represented by parabolas over a limited range of doses. So it is desirable to have a simple method for estimating the potency of an unknown, relative to a standard preparation from two parabolic response curves.
Symmetrical 2k-point designs with k=3, 4, and 5 for obtaining a point estimate and an approximate interval estimate of the relative potency of two parallel parabolas are described here. The method of validity tests is also included. A numerical example of bioassay of dimethylcurine methylchloride is given with calculation procedure fully illustrated. The methods remain applicable even when the integer k (>5) is arbitrary.

Article Options

Download Citation

Cited times in Scopus