The binding constants and structural components of 200 drugs and enzyme inhibitors have been used to calculate the average binding energies of 10 common functional groups. As expected, charged groups bind more strongly than polar groups, which in turn bind more tightly than nonpolar groups. The derived intrinsic binding energies (in kcal/mol) are (i) charged groups, C02", 8.2; P041 2', 10.0; N+, 11.5;(ii) polar groups, N, 1.2; OH, 2.5; CO, 3.4; O or S ethers, 1.1; halogens, 1.3;(iii) nonpolar groups, C (sp2), 0.7; C (sp3), 0.8. These values may be used to determine the goodness of fit of a drug to its receptor. This is done by comparing the observed binding constant to the average binding energy calculatedby summing the intrinsic binding energies of the component groups and then subtracting two entropy related terms (14 kcal/mol for the loss of overall rotational and translational entropy and 0.7 kcal/mol for each degree of conformational freedom). Drugs that match their receptors exceptionally well have a measured binding energy that substantially exceeds this calculated average value—examplesinclude diazepam and biotin. Conversely, if the observed binding energy is very much less than the calculatedaverage value, then the drug apparently matches its receptor less well than average. Examples of this type include methotrexateand buprenorphine.

The experimentally observed binding of a drug to its receptor provides a measure of the total interaction be-tween the two molecules but generally tells us little or nothing of the three-dimensional quality of their interaction. Is the match as good as that ofhand and glove, or more like that of square peg and round hole? Morphine, for example, binds at 5 nM and is often regarded as a prototype drug for analgesic activity, but how well does it match its receptor? Butaclamol, being comparatively rigid, is widely used as a model for dopamine receptor antagonist activity, but how many of its functional groups actually interact with the receptor? Peptidessuch as the enkephalins are thought to bind to their receptors via several amino acids, but how many are actually required to account for the observed binding? Oxalate anion has an inhibition constant, Kh for trans carboxylase1* of 1.8 X 1CT6 M, while methotrexate has a Kx of 10 “n for dihydrofolate reductase, 2 but which represents the best match to the corresponding active site? In conformationally flexible molecules, we know that the lowest energy conformation is not necessarily the biologically active form, since part of the drug-receptor binding may be used to perturb the drug conformation, but how much above the global energy