Klotz I M, Hunston D L
J Biol Chem. 1984 Aug 25;259(16):10060-2.
In the basic life sciences the term "model" implies a physical, chemical, or molecular construct that provides a representation for the interpretation of experimental observations. To the statistician, however, a model is a mathematical expression for correlating data, which may or may not have roots in a molecular picture. With regard to ligand-receptor interactions, the mathematical model used plays a crucial role in extrapolations of binding measurements. Regardless of the statistical goodness of fit of data to an equation, the relationships of the parameters of a mathematical formalism to the molecular features of ligand-receptor complexes are generally very complex. Oversimplified interpretations of the molecular significance of the constants derived from binding measurements are unwarranted, unless one has independent information from molecular probes.