« Happy Fourth of July |
Main
| How Do You Know When to Stop? »
July 6, 2004
Lighting Out for the Frontier
Posted by Derek
Next week I'll be on vacation, and this week I find myself flogging the chemistry databases. We're looking at a new series of compounds with some interesting activity, and they're easy to make. That's the problem.
They're so easy to make that everyone has made them. That means trouble, because there's so much prior art (and so many potentially interfering claims) that we'd have a difficult time getting a patent position. There are quite a few popular structural classes with this problem: piperazines, indoles, and imidazoles have been trampled flat, for example, and try getting (for example) a claim in some of the benzofuran compound space that Lilly has staked out. Not fun.
So we're looking to change the inner parts of the structure in some non-obvious way and thus stake out a position with some elbow room in it. It's not easy - for one thing, things are to the point that some of the obvious paths to elbow room have already been worn out a little. And you don't want to come up with new ones that are going to take fourteen steps to make, either. That's because it's unclear whether the interesting activity that got us here will persist once we start chopping and rearranging.
Often it doesn't. You feel like a real idiot - even if you aren't - when you spend a lot of time on a series like that and none of the compounds are active. And you feel as if you've wasted our time, because you have. "Fail quickly" is one of the basic mottos of medicinal chemistry.
Comments (3)
+ TrackBacks (0) | Category: Life in the Drug Labs | Patents and IP
- RELATED ENTRIES
- › Shire's Replagal Problems: An Inside Look?
- › Bungled Structure, And How
- › Roche Closes Nutley, Once Its US R&D Home
- › The Next Five Years in the Drug Industry
- › A Kinase Inhibitor Learns Something New
- › The Good Ol' Friedel-Crafts
- › Merck's Madagascar Marketing Mess
- › Scientific Literacy: Where Do You Stop?
1. The Novice Chemist on July 7, 2004 2:00 PM writes...
Permalink to CommentIs "fail quickly" a difficult thing to learn in industry? Grad school is really a perseverance thing and the concept seems odd to me. (Then again, it doesn't. If a synthetic route isn't going to work, you don't want to keep working on it.)
2. qetzal on July 7, 2004 5:28 PM writes...
In my experience, the concept of "fail quickly" is not the tricky bit. It's obvious that there's no point working on something that cannot succeed. The hard part is, how do you know when something won't work?
Presumably, all projects start from a belief that they might work. (OK, maybe that's optimistic, but can we assume that for now? Thanks!)
Over time, one accumulates data, which commonly suggests that the project isn't likely to work after all. But I find that one can almost always come up with alternate ideas, approaches, and explanations why it can still work, even in the face of discouraging results. And I don't just mean in the self-serving, please-don't-kill-my-pet-project sense.
It can be really hard to decide when you've tried all the reasonable avenues, and any residual chance of success is too small to justify further effort. And it doesn't help when we hear the stories of such-and-such a drug that was killed three different times, only to ultimately rise from the ashes, get approved, and make good money.
The art of killing projects is difficult to master, IMHO.
Permalink to Comment3. qetzal on July 7, 2004 5:45 PM writes...
Novice Chemist's question suggests an interesting blog topic for Derek, if he's reading.
What things are generally not taught to science PhD's, but would be really valuable (especially for people like us who end up in industry, be it big pharma, biotech, whatever)?
Some examples from my experience:
Budgeting & accounting. At least in small biotech, as soon as you make it to a middle mgmt level, you'll likely be asked to prepare budgets for your dept/project, and keep track of your spending. I had no clue how to do that the first time around. And judging from what I've seen in biotech, it's a very common problem.
Leadership & management. How many science PhD's end up having to manage staff? A large fraction. And that applies academics as well as industrial types. When I first had people reporting to me, I had zero training in how to manage and lead their efforts. Even now, almost all my training has been on-the-job experience, supplemented with the occasional half-day course.
And of course, it's not just that these things aren't taught in grad school. Science types rarely take such courses as undergrads either.
Derek, if you read this, maybe you'd like to expand on the theme. Seems like it would make an interesting installment to your "Industry vs Academia" series.
Permalink to Comment