Comparison of experimental estimates and model predictions of protein crystal nucleation rates

Estimates of protein crystal nucleation rates determined from three independent sets of experimental observations are compared for internal consistency. Large discrepancies emerge. A qualitative understanding of these discrepancies is gained from an analysis of the experimental techniques used within the framework of extant models for nucleation rates. Quantitatively, however, these models fail to predict nucleation rates of protein crystals. Possible origins of this failure lie in the approximate descriptions of the interaction potentials between protein molecules and the processes of aggregation and dissociation of single molecules on cluster surfaces that lead to crystal nucleation.