Efficient Estimation of Protein Binding Constants

Assays of cyclic AMP typically employ a Scatchard‐like data analysis to estimate the total binding protein R_T and the dissociation constant K_d. The linear regression procedure employed is statistically ad hoc because the response variable (bound counts) occurs on both sides of the Scatchard equation. The alternative analysis presented here is derived from the cyclic AMP binding reaction. The resulting data analysis, employing non‐linear least squares, results in estimators that on the basis of Monte Carlo studies have less bias and greater precision than the Scatchard method.