Thank you for your comments. What exactly is being measured in these experiments is indeed important. The on rates are reported in units of M^(-1)s^(-1), which was by far the most common convention we found in the literature. Very few, if any, of the rates were measured using BLI, although many were determined by SPR. It is our understanding that the resonance angle is sensitive to any material in the nanometers above the surface, so presumably the encounter complex is being detected. We are told that the response is linear with mass concentration, which suggest to us that SPR can't be used to tell the difference between a diffuse or narrow ensemble. That said, neither of us are experts. As for the relevance to our work, if we exclude extreme examples like the mutant EIN-HPr complex studied by Tang et. al (Nature, 444, 7117:383-6.), then the encounter complex is only sparsely populated relative to the bound and unbound states (FEBS Lett., 583, 7:1060-6.). It follows that the microscopic rates of dissociation/association from the encounter complex are fast compared to the macroscopic rate constants. Once we account for experimental uncertainties, unmodelled pH, buffer, ionic strength and temperature effects, as well as the errors associated with the conformational sampling and energy calculation methods, then the influence of the experimental 'bound' including the twilight zone between dissociated and tightly bound is likely absorbed into the noise.

Iain Moal and Paul Bates