Your question is a good one, though it's somewhat unclear. I think what you're asking, in essence, is whether one needs quantum mechanics to accurately model the activity of proteins, or whether classical physics is sufficient. If one takes the proteins as already existing, the answer, in the vast majority of cases, is that classical mechanics is sufficient. In fact, the standard "molecular dynamics" methods of computational simulations of proteins binding to one another, changing shape, etc., are all classical in nature. In a bit more detail:
The chemical bonds that hold atoms together in molecules, such as proteins, are inherently quantum mechanical in nature; classical physics is insufficient to explain them. But, once one "has" a molecule in hand and wants to think of this object as a whole, one needn't (in most cases) worry about the substructure. Very roughly, one can get a sense of how important quantum mechanics is by thinking about the de Broglie wavelength of an object, compared to its size. For a protein with ~kT~$kT$ of energy, the thermal de Broglie wavelength is ~ 10^{-12}$ 10^{-12}$ m, much smaller than the size of the protein; its wave-like character is largely irrelevant. Modeling the protein, therefore, is a fine job for classical physics. As mentioned, this is, in fact, what is generally done, quite successfully.
There are exceptions, of course, but these tend to involve situations in which the motions of electrons are much more important, such as excitations in photosynthesis. (Electrons have much smaller mass, and so much greater wavelengths.)
