[Dav]

how to explain the nature of chemical's pharmacological activity is governed by its chemical properties or its structure

[orgopete]

This question seems like looking at the back end of a horse. If we asked this question about aspirin, I would first ask you what the cyclic oxidase enzyme does. Since aspirin inhibits that enzyme, is the activity of aspirin governed by its chemical properties or its structure (as if they are different).

This is not my favorite example as any structural similarity between the substrate and inhibitor is either absent or not obvious. Perhaps someone might offer up some examples showing the structures of the substrates and their drugs. Perhaps a drug like methotrexate and tetrahydrofolate would suffice.

Then, if you are like me, you will say to yourself, "Hey, it looks like the substrate." Drugs are like the sidewalk Rolex watches.

It isn't quite that simple. The "looks like" is what leads to the notion of "lock and key" except enzymes are driven by energy or forces. If you have a "hard working" enzyme, you can find the inhibitors can vary quite far from the substrate structure. I define a "hard working" enzyme as one that catalyzes a reaction with a high activation energy. Enzyme work by reducing the activation energy of a reaction and thus increase the reaction rate. Thus a "hard working" enzyme will have a high affinity for the intermediate or transition state of its substrate. That is the concept that Pauling had suggested. In practice, you may find enzymes are rather more promiscuous than the mimicry one might find with some drugs and their substrates.

[zaphraud]

Because many neurotransmitters are small molecules, and because the "blood brain barrier" provides a simplified environment, it is possible to see clear and strong relationships between chemical properties and drug effects. Sure, a strictly computational approach (QSAR) seems to seriously leave something to be desired, but the relationships are at least visible, and if not interpreted quantitatively but rather by a human being, actually make a whole lot of sense.

You do not need to refer to literature describing restricted compounds to see these relationships, either. It is also easily seen in drugs used in psychopharmacology, or even just studies done on the various undesirable properties of antihistamines, such as sedation.

The effect of drugs in the entire body is harder to predict because there are vastly more systems involved, so if the primary activity being studied isn't CNS-related, its might be harder to learn from existing examples, I think.