The question is actually more interesting than it sounds since it really goes to the heart of what drug discovery is about. In some sense the question does not have an answer since drug discovery is one of the most multidisciplinary research endeavors humans have created, so it's pretty much impossible for any one scientist to discover any drug except by sheer accident. And yet some of us ventured interesting answers to Lisa's question: My guess was 10%, although I was thinking about someone who was part of a whole team that invented the drug. I would probably degrade that guess to 5% on second thoughts. Derek's guess was "less than 1%" and he was referring to the probability of someone actually synthesizing the drug molecule with their own hands. Chemjobber ventured a similar number.
But these numbers only make the question more complicated. For instance, for better or worse, you may be - and often are - a cog in the wheel of a whole machinery of chemists and other scientists working on a drug project. Sometimes by sheer luck you just happen to make the winning molecule that turns out to have the right properties: any other chemist would have been equally competent to make the same molecule. Does that make you the principal inventor? Most people would say no and they would be right, since such a claim would ignore the massive amounts of background knowledge and thinking generated by your entire team that generously littered the path leading to that particular compound. I would say that at the very least, the lead medicinal chemist who heads the team, coordinates different results and ideas, assigns tasks and holds the big picture in his or her mind is equally responsible for the discovery, as long as they are not playing a purely managerial role.
Judging a chemist as one who "invented" a drug would presumably involve finding him or her cited on one or more key inventions related to the drug. Unfortunately the question about chemists discovering particular drugs being on key patents also highlights the features of what I consider to be an outdated patent system which still talks about "inventions" rather than "ideas". The overhauling of this system is especially pertinent in an age where design in chemistry has become much more important than synthesis. Granted that inventions are more tangible and easy to quantify, but without the set of interdisciplinary ideas that contribute to the discovery of a drug there would be no inventions to begin with. And a vast multiplicity of scientific talent and not just medicinal chemistry is responsible for initiating, assimilating and exchanging this set of ideas.
Yet over the years the patent system has become biased toward chemists because their contribution to "inventions" is easiest to measure. It's worth remembering in this context that the original patent system arose from the system of industrial research in late 19th century Germany in which chemists were the leading, if not the only players. Almost nothing about the rational design of molecules was known then, so all the glory and money went to practical chemists who tinkered with molecular combinations and came up with the right one by pure trial and error. The current patent system has been more or less directly inherited from that rather biased framework. Thus you will seldom find biologists, pharmacologists, crystallographers, toxicologists or molecular modelers cited as frequently on patents as medicinal chemists, and yet their contributions are as or sometimes even more important. For instance, it's increasingly appreciated that the biggest obstacle to the discovery of a new drug in a novel disease area is the lack of reproducible and realistic biology; by this token biologists should be the most important contributors to the discovery of a new drug, and often they are. Yet the problem is that these contributions are also harder to classify as inventions. I think there has been a much bigger recognition of constructs such as molecular models, assays and model organisms on patents in recent times, but chemists who make things with their own hands still have the upper hand.
So what is the probability of a chemist discovering or inventing a drug in his or her career? Whatever the probability is, most people would agree that it is very low because of the role that sheer luck plays in developing a new drug. Some of the most promising candidates in the process fall out at the last minute; conversely some dark horses can emerge to be winners out of the blue. But the path to these winners and losers is seldom the handiwork of a single scientist. Sometimes certain individuals rightly take the stage, and they should be duly recognized. But we should never forget those in the wings, down in the orchestra and up in the light box without whom these individuals would never be standing up there and taking a bow.
Image link