About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Vivus and Qsymia: The Oddest Drug Approval | Main | Quick Links »

July 18, 2012

The Best Rings to Put in Your Molecules?

Email This Entry

Posted by Derek

Here's a paper from some folks at GlaxoSmithKline on what kinds of rings seem to have the best chances as parts of a drug structure. They're looking at replacements for plain old aryl rings, of which there are often too many. Pulling data out of the GSK corporate collection, they find that the most common heteroaromatic rings are pyridine, pyrazole, and pyrimidine - together, those are about half the data set. (The least common, in case you're wondering, are 1,3,5-triazine, 1,3,4-oxadiazole, and tetrazole). In marketed drugs, though, pyridine is more of a clear winner, and both pyrrole and imidazole make the top of the charts as well.

When they checked the aqueous solubility of all these compounds, the 1,2,4-triazoles came out on top, and the 1,3,5-triazines were at the bottom, which sounds about right. Other soluble heterocycles included 1,3,4-oxadizole and pyridazine, and other bricks were thiazole and thiophene (not that that last one really counts as a heterocycle in my book). Update: I've revised my thoughts on that! Now, you might look at these and say "Sure, and you could have saved yourself the trouble by just looking at the logD values - don't they line up?" They do, for the most part, but it turns out that the triazines are unusually bad for their logDs, while the five-membered rings with adjacent nitrogens (all of 'em) were unusually good.

The next thing the team looked at was binding to human serum albumin. The 1,3,4-thiadiazoles emerged as the losers here, with by far the most protein binding, followed by thiazoles and 1,2,4-oxadiazoles. Imidazoles had the least, by a good margin, followed by pyrazine and pyridazine. Those last two were better than expected compared to their logD values.

And the last big category was CYP450 inhibition. Here, thiophene, tetrazole, and 1,2,3-triazole were the bad guys, and pyridazine, 1,3,4-thiadizole, and pyrazine (and a few others) were relatively clean. The people at AstraZeneca have published a similar analysis, and the two data sets agree pretty well, with the exception of oxazole and tetrazole. The AZ oxazoles all had open positions next to the ring nitrogen, which seems to have opened them up to metabolism, but the difference in tetrazoles (AZ good, GSK bad) is harder to explain.

The take-home? Pyridazine, pyrazine, imidazole and pyrazole look like the winners from an overall "developability" score. Thiophene brings up the rear, but since I still think that one shouldn't count update (it's a benzene in disguise), the ones to worry about are then thiazole, 1,2,3-triazole, and tetrazole (that last one with an asterisk, due to the CYP data discrepancy).

The paper tries to do the same analysis with heteroaliphatic rings, but the authors admit that they had a much smaller data set to work with, so the conclusions aren't as strong. There was also a higher correlation with plain ol' logD values across all three categories (not as many surprises). The winners turned out to be piperidine NH and morpholine N-alkyl, with imidazoline and piperidine N-alkyl right behind. The losers? Piperidine N-sulfonamide, followed by pyrrolidine N-sulfonamide, and then 1,3-thiazolidine. (Sulfonamides continue to live up - or down - to their reputation as Bad News).

There are, naturally, limitations to this sort of thing. Ceteris paribus is a mighty difficult state of affairs to achieve in medicinal chemistry, and other factors can rearrange things quickly. But if you're just starting out in an SAR series, it sounds like you might wand to give the pyrazines and pyridazines a look.

Comments (14) + TrackBacks (0) | Category: Life in the Drug Labs


1. Anonymous on July 18, 2012 11:29 AM writes...

Derek, the two links point to the same place, not different papers.

Permalink to Comment

2. Will on July 18, 2012 11:53 AM writes...

Given the success of the sartan class of angiotensin antagonists, I would not be too quick to dismiss the tetrazole

Permalink to Comment

3. Derek Lowe on July 18, 2012 12:11 PM writes...

Fixed the second link - thanks!

Permalink to Comment

4. Henning Makholm on July 18, 2012 12:26 PM writes...

Um, not a chemist here, but what did the poor thiophene do to you not to be considered a heterocycle? Does it spontaneously rearrange to put the sulphur outside the ring or what?

Permalink to Comment

5. milkshake on July 18, 2012 2:31 PM writes...

I love benzfurazanes - easy preparation from nitroanilines, good patent position, good metabolic stability, more polar but more el deficient than phenyl, act as a decent isosteric replacement of a nitrophenyl group. Mildly explosive but there are several process papers in OPR&D on benzfurazane drug candidates.

Permalink to Comment

6. sean on July 18, 2012 4:10 PM writes...

Will this be more useful than Lipinski's rule-of-thumb (I mean 5)?

Permalink to Comment

7. David Formerly Known as a Chemist on July 18, 2012 4:19 PM writes...

Sorry Derek, but I've personally seen multiple examples in several series where switching from thiophene to phenyl led to several log orders of activity loss, or led to greatly increased plasma protein binding, or some other drastic physicochemical change. I can't agree that a thiophene is a benzene in disguise (if so, it's a very poor disguise).

Permalink to Comment

8. milkshake on July 18, 2012 7:43 PM writes...

The two important differences between thiophene and benzene are 1) easy thiophene oxidative metabolic ring opening - this could be curse or unexpected blessing, see for example clopidrogel 2) Thiophene sulfur (and also thiazole sulfur) bears a significant partial positive charge and therefore likes to coordinate to lone el pairs of carbonyl oxygens and pyridine/pyrimidine nitrogens. This can contribute to stronger binding in active site but more importantly in drugs like kinase inhibitors (that contain several heterocycles) the S...O=C or S...N interaction locks in one particular planar conformation on the part molecule.

Permalink to Comment

9. Anonymous on July 18, 2012 8:02 PM writes...

Call me crazy, but I thought the point of tetrazoles and sulfonamides was more to mimic acidic moieties?

Any words on lactams (lactones) or cyclic (thio)ureas?

Permalink to Comment

10. MoMo on July 19, 2012 9:41 AM writes...

Pyrazines and their lone pair N atoms give them special properties-they act like water-soluble benzenes.

Good paper for paper chemists!

Permalink to Comment

11. Ever Decreasing Circles on July 19, 2012 12:28 PM writes...

Have I got this right?

Take a list of marketed drugs, find a descriptor (any descriptor) that distinguishes them from your in-house/patented compound collection and publish a paper announcing "Four legs good, two legs bad".

What does this this say about the confidence and vigour of small molecule Drug Discovery?

One thing is certain: any more of these rules and the only structures left as "acceptable" to develop will be those already on the market.

Permalink to Comment

12. ThirdWheel on July 19, 2012 12:39 PM writes...

Repeat after me, correlation is not causation, correlation is not causation, correlation is not causation....

Permalink to Comment

13. TX raven on July 20, 2012 12:47 AM writes...

@ 11 and 12:

You hit the nail in the head.

Permalink to Comment

14. Joe Lyon on July 20, 2012 8:26 AM writes...

Derek, have you got the right AstraZeneca paper? -there's one in Medchemcomm published last week that seems more relevant:

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry