Field of Science

What is the probability of a chemist discovering a drug? Is that the right question?

Over on Twitter there has been a lively discussion sparked by a question from C&EN's Lisa Jarvis: What is the probability that a medicinal chemist will discover a marketed drug in his or her career? I have some interest in this question since I happened to directly work with two world-class medicinal chemists who invented/discovered bestselling drugs.

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


  1. Under US patent law, inventions ARE ideas and inventors are the one(s) who articulate the idea for the first time or contribute to it in its final form. The same rules apply to all patents, not just drug patents. No, the patent law does not favor chemists, and certainly not because it is easier to measure the contribution of chemists than other scientists. For a composition of matter patent, the inventor(s) may or may not include the person who made the compound or the lead chemist, and I can tell you that sorting that out is not easy. Drug patents include composition of matter and other types, and clearly the former are going to be mostly chemists. The others, not so much. Invention should never be confused with discovery. A drug is "discovered" by a team but invented (usually) by a much smaller group. The two terms should not be used interchangeably. Sorry, but you're just way off base on this.

  2. According my understanding, "articulate the idea" is not sufficient - it also need enablement by this person. We have enough peopble with brilliant ideas, especially afterwards and oftern in opposite direction over the time span in a team.

  3. You have to fully articulate the idea, including how it would be reduced to practice. For instance, for a drug compound, you need to say, among other things, what disease it would be useful for treating (the utility) and how it could be made (the process). The "idea" also needs to be non-obvious (often difficult to define) based on the prior art, so just proposing a new analog does not necessarily qualify. You do not necessarily need to have made the compound, but you need to say how it could be made. The patent can be invalidated if it turns out that this method does not work, although "work" is defined very broadly. My point, however, was to emphasize the often misunderstood difference between inventorship (which has a legal definition) versus discovery, which has a scientific and business definition. I also found the statement that the patent system favors synthetic chemists baseless. Biologists, analytical chemists, physicians, drug metabolism experts, formulation scientists, and more can and are listed as inventors on drug company patents that, for instance, cover the discovery of new analytical or assay procedures, new biomoarkers, new active metabolites, and more, all of which contribute to the "discovery" of a drug. Even ***molecular modelers** are listed as inventors on compound patents when they are responsible for proposing a new type of analog (slight friendly dig, there). Granted, companies tend to file many more composition of matter patents than anything else, and granted, there are many issues with the patent system. We should remember that it was not designed as nor should it be viewed as a vehicle for recognition of achievements. There is, for instance, no such thing as a "lead inventor" or "corresponding inventor" as there is in scientific publications. You either are or you aren't an inventor, and not everyone qualifies for every patent, no matter how vital their contribution to the discovery process. In my experience, when it comes time to hand out the awards for a great "discovery", everyone lines up for their prize, and most everyone who has contributed in some way is recognized, just not with their name listed as an inventor.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS