Radioimmunoassay data processing, ch. 2

The problems of radioimmunoassay data processing with a fairly big computer are investigated in detail. The logit transformation, introduced by Rodbard et al. in 1968, and the related mathematical operations are presented. From logit-ln plots the confidence limits for the calculation of the antigen concentration can be predicted as well as the minimum detectable dose (i.e. the sensitivity of the assay). The Scatchard plot is used to calculate the equilibrium constant of the antigen-antibody coupling reaction and the binding capacity. A 4-parameter logistic curve fitting method is developed and combined with a method to carry out a multi-component Scatchard analysis. Both a weighted and an unweighted linear regression procedure are evaluated for use in calculating the plasma renin activity from the angiotensin I amounts determined by radioimmunoassay. Improvements for Rodbard's computer programmes are proposed