Statistical assumptions underlying the fitting of the Michaelis-Menten equation

An experiment was carried out to test the various assumptions usually made when evaluating statistical procedures for estimating the parameters of the Michaelis Menten equation, which describes enzyme-catalyzed reactions. The usual assumption of normality is not strongly supported, but is probably not too unreasonable. We study the variation in experimental results and, in consequence, a more complex model is proposed, which incorporates extra components of variation associated with substrate levels and diff erent days. The model is fitted using the EM algorithm.