There's a possible new area for drug discovery that's coming from a very unexpected source: enzymes that don't do anything. About ten years ago, when the human genome was getting its first good combing-through, one of the first enzyme categories to get the full treatment were the kinases. But about ten per cent of them, on closer inspection, seemed to lack one or more key catalytic residues, leaving them with no known way to be active. They were dubbed (with much puzzlement) "pseudokinases", with their functions, if any, unknown.
As time went on and sequences piled up, the same situation was found for a number of other enzyme categories. One family in particular, the sulfotransferases, seems to have at least half of it putative members inactivated, which doesn't make a lot of sense, because these things also seem to be under selection pressure. So they're doing something, but what?
Answer are starting to be filled in. Here's a paper from last year, on some of the possibilities, and this article from Science is an excellent survey of the field. It turns out that many of these seem to have a regulatory function, often on their enzymatically active relations. Some of these pseudoenzymes retain the ability to bind their original substrates, and those events may also have a regulatory function in their downstream protein interactions. So these things may be a whole class of drug targets that we haven't screened for - and in fact may be a set of proteins that we're already hitting with some of our ligands, but with no idea that we're doing so. I doubt if anyone in drug discovery has ever bothered counterscreening against any of them, but it looks like that should change. Update: I stand corrected. See the comment thread for more.
This illustrates a few principles worth keeping in mind: first, that if something is under selection pressure, it surely has a function, even if you can't figure out how or why. (A corollary is that if some sequence doesn't seem to be under such constraints, it probably doesn't have much of a function at all, but as those links show, this is a contentious topic). Next, we should always keep in mind that we don't really know as much about cell biology as we think we do; there are lots of surprises and overlooked things waiting for us. And finally, any of those that appear to have (or retain) small-molecule binding sites are very much worth the attention of medicinal chemists, because so many other possible targets have nothing of the kind, and are a lot harder to deal with.