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

« Ten Tons of Sodium And Just One Lake | Main | Scaling Up a Strange Dinitro Compound (And Others) »

February 21, 2012

Rational Drug Design. Hmm.

Email This Entry

Posted by Derek

Here's a huge review that goes over most everything you may have wanted to know about what's called "rational drug design". The authors are especially addressing selectivity, but that's a broad enough topic to cover all the important features. (If you can't access the paper, here's a key graphic from it).

"Rational", it should be understood, generally tends to mean "computationally modeled" in the world of drug discovery. And that's certainly how this review is pitched. I'm of two minds - at least - about the whole area (a personal bias that has made for some lively discussions over the years). Some of those discussions have taken place between my own ears as well, because I'm still not sure that all my opinions about computational drug design are self-consistent.

On the one hand, drug potency is a physical act which is mediated by physical laws. Computing the change in free energy during such a process should be feasible. But it turns out to be rather difficult - proteins flex and bonds rotate, water molecules assist and interfere, electrostatic charges help and hinder, hydrogen bonds are vital (and hard to model), and a dozen other sorts of interactions between clouds of electrons weigh in as well. Never forget, too, that free energy changes have an entropy component, and that's not trivial to model, either. I keep wondering if the error bars of the various assumptions and approximations don't end up swamping out the small changes that we're interested in predicting.

But, on that other hand, there are certainly cases where modeling has helped out a great deal. A cynic would say that we've been sure to hear about those, while the cases where it had no impact at all (or did actual harm) don't make the journals very often. It can't be denied, though, that modeling really has been (at times) the tool for the job. It would be interesting to know if the frequency of that happening has been increasing over time, as our tools get better.

Because on the third hand, it's been a poor bet to go against the relentless computational tide over the last few decades. You'd have to think that sheer computing power will end up making molecular modeling ever more capable and useful, as we learn more about what we're doing. Mind you, there were people back in the mid-1980s who thought we'd already reached that point. I'm not saying that they were the best-informed people at that time, but they certainly did exist. I wonder sometimes what it would have been like, to show people in 1985 what the state of rational drug design would be like in 2012. Would they be excited, or vaguely disappointed?

And then there's that word "rational". I think that its adoption might have been the best advertising that the field's ever achieved, because it makes everything else seem irrational (or at least arational) by default. I mean, do you just wanna make compounds, or do you want to think about what you're doing? I also wonder what might have changed if that phrase had never been adopted - perhaps expectations wouldn't have gotten out of hand in the computational field's early days, but it might not have received the attention (and money) that it did, either. . .

Comments (35) + TrackBacks (0) | Category: In Silico


1. milkshake on February 21, 2012 12:19 PM writes...

Derek, it is not all just molecular modeling (which I agree is not very helpful when used alone without solid structural information). I think rational drug design means "target-driven drug discovery project that uses small molecules whose design is based on SAR insights derived from protein crystal structure and is typically informed by computer modeling. It is a process of chemical modification that is done in a methodical fashion and conforms to precepts like rule of 5 and polar surface area limits"

Permalink to Comment

2. CR on February 21, 2012 12:47 PM writes...

Actually, Milkshake, I would expand on the definition of rational drug design. I don't think you need protein crystal structures or computer modeling at all to be "rational". Working in the GPCR space where there is no structural information (at least for my project), I perform "rational" drug design everyday. I look at the previous round of SAR and "rationally" design compounds.

I think the Bogarting of the term "rational drug design" is for molecular modelers trying to feel they are part of the process (and not an afterthought). What I've always found enjoyable are the modelers that want to do "rational drug design" and then propose a molecule that could never be made (what, you can't make an O-O-O bond???).

Permalink to Comment

3. Pharmaheretic on February 21, 2012 12:56 PM writes...

It seems that beta-secretase inhibitors also have adverse effects on memory.

Permalink to Comment

4. Curious Wavefunction on February 21, 2012 1:38 PM writes...

I agree; "rational" drug design is by no means limited to computational methods. And as someone above hinted, rational drug design can be both target as well as ligand-based, although it has most often been defined in a structure-based context.

As for its virtue, it's clear that drug design is too complex a process to succumb to our definitions of rationality. Our understanding of biological systems is just not good enough to make rational predictions about their interactions with small molecules. What we need (and what has always worked) is a wily mix of rationality, serendipity and intuition. Figuring out the exact composition of that mix has always been the challenge, and will remain so for the conceivable future.

I always say that rational drug design should be called " retrospect".

Permalink to Comment

5. Jose on February 21, 2012 1:46 PM writes...

If rational drug design existed, the top 2-3 people in the field would be consultants with weekly contracts in the tens of millions. Drug design has been, and still is, dumb-luck + trial and error.

Permalink to Comment

6. haha on February 21, 2012 1:50 PM writes...

Is there any irrational drug design? By definition, there must be a rationale behind all design although it does not mean that the rationale is always correct.

Permalink to Comment

7. entropyGain on February 21, 2012 1:52 PM writes...

As a recovering computational chemist who was very active at the dawn of the commercial computational chemistry software industry (early/mid 1980s), I am very disappointed with the state of innovation in the field over the last 30 years. Other than more cpu power and color graphics, there has been little new development in the way we use computers to assist drug discovery. FoldIt is probably the most innovative new concept I can think of. But all the time wasted on "diversity" is appalling. Docking is almost unchanged. What ever happend to Sculpt? (that was cool).

The sofware companies spent WAY too much time figuring out how to implement licensing methods, porting to the latest software architecture or hardware, or wrapping a GUI around some academic software. Far too little time was spent sitting with chemists/biologists/project teams and figuring out what would really help.

Permalink to Comment

8. Anonymous on February 21, 2012 2:15 PM writes...

entropyGain - I agree that the original path of comutational chemistry from the 1980's to ~2005 has leveled off. But SAR and protein folding apart, does "disease modeling" not refect where more computational investment (if not advancements) have been made?

Permalink to Comment

9. Curious Wavefunction on February 21, 2012 2:27 PM writes...

entropyGain, I would argue that docking is far better than where it was fifteen years or even a decade ago. It still has a long way to go, especially in the prediction of binding energies, but being able to predict the correct protein-bound ligand structure accurate to within 2 A RMSD of the crystal structure in about 30% of the cases is no small achievement, especially when you are starting from a random static structure of the unliganded protein. In addition one could argue that we now have a much better handle on the strengths and limitations of the technique and have at least tried to realistically incorporate the basic physics of the process into our algorithms.

Permalink to Comment

10. Drug Developer on February 21, 2012 2:31 PM writes...

I'll go even further with broadening the definition: "The opposite of combinatorial chemistry"...

Permalink to Comment

11. CMCguy on February 21, 2012 3:13 PM writes...

#10 DD I take umbridge at perceived slight for stereotyping of combichem as totally irrational. It was spawned by basic reasoning had all these wonderful biologic assays that could be readily adapted to HTS so if one then only threw enough molecules at them that would get many new leads, which in turn could be further narrowed by making directed libraries. Behind that most of the designs or scaffolds often had consideration chemistry rationales behind them including med chem experience, however was typically predominately driven by how many variants can be acquired. Problems occurred because all the over-hype with flaws in the underlying motivations that did not translate to real promise (other than as a possible supporting tool of certain utility) but not sure combichem fared any worse than all the computer modeling efforts for generation of candidates and successful drugs to market. It can be part of #4 CW's "wily mix" with the challenge of when to use effectively.

Permalink to Comment

12. Boghog on February 21, 2012 3:18 PM writes...

My favorite definition:

"Rational drug design is the inventive process of finding new medications based on the knowledge of a biological target." Rational drug design (Wikipedia)

"Rational drug design" is often equated with "computer-aided drug design", but it is certainly possible to design a drug without the aid of computers.

Permalink to Comment

13. Hap on February 21, 2012 4:16 PM writes...

It's more damaging to think that I know what I'm doing when I actually don't. Sometimes the biology is not well-known enough to design a ligand for a target because you don't know what the target is. It seems that "rational drug design" implies knowledge of all the variables in biology and drug design, knowledge that no one really seems to have.

HTS is probably "irrational drug design", but considering that the limitations of our knowledge may not be clear (or known), it probably isn't as irrational as it seems. It's better than drawing "Here be monsters" on the map and assuming we know what's out there - it's pretty clear that we don't know what's out there, let alone how it all works.

Permalink to Comment

14. RD on February 21, 2012 6:14 PM writes...

Haven't read the paper yet but having done it for years, there are times when comp chemists overmodel. Yeah, I think there is a limit to precision and the time it takes to determine something may take away from more gross observations that would be more helpful. The projects that I think benefit from modeling the most are the ones where there is good structural biology support. Sometimes, you just have to stare at the binding modes until something jumps out at you that you hadn't noticed before.
Then again, chemists can be so hostile to modeling that you might as well not have a modeler on the project at all. The chemists spend several years making compounds before they find out that the modeler could have told them a couple of years earlier that a certain approach wouldn't work. So, a little more cooperation on the synthetic side would be greatly appreciated.

Permalink to Comment

15. Bombasticus on February 21, 2012 7:04 PM writes...

Ever increasing computational power and lower energy per calculation means in silico is the way forward. The big problem is one perpendicular in a sense to both computing and chemistry, how to model accurately enough to be useful, which is ultimately a blend of physics, and programming. Chemistry then becomes interrogating and manipulating the model in the right way. Nobody can rationally argue that the hardware capability doesn't already exist or at the very least isn't very close to existing, nobody can argue that these problems aren't mathematically soluble, after all reality is a mathematical construct, although our best descriptions of it may lack some nuance. All that is at issue is time, lots of it or... Not exactly a state of the union on the state of 'rational' or more correctly computational drug design but they are the unalterable facts. Chemistry isn't going anywhere, drug discovery will still be in the realm the chemist calls home, he or she will just have much, much better tools. Instinct and intuition on the part of the chemist will become, if anything, even more important to leverage such tools. When? I'll just get my tarrot cards...

On a side note where did you get the third hand derek and does it come in handy in the lab?

Permalink to Comment

16. ScientistSailor on February 21, 2012 8:21 PM writes...

#6 HaHa...There is plenty of irrational drug design. Open a BMCL recently? Half those papers are simply testing compounds to see if they kill cells. That's irrational...

Permalink to Comment

18. OldLabRat on February 21, 2012 9:16 PM writes...

Perhaps calling it non-random would be a more effective description than rational. Part of the problem is that many of the comp chemists that also contribute to software development have a biased world-view that doesn't admit med chem experience being rational. This despite the empiricism of many computational models.

The other thought I had after reading the article: it really helps demonstrate how hard drug discovery is. All the considerations the authors cite just to get potency. Now add demonstrating oral PK/PD, low toxicity, reasonable cost of goods, manageable adverse events and then one understands the challenge.

Permalink to Comment

19. MoMo on February 21, 2012 10:18 PM writes...

Stop it NOW! All of you! Especially the Brains that wrote that paper we are all gushing over!

Rational Drug Design has been around since aliens landed in Roswell, NM- the late 1940-50's with E J Ariens, the undisputed father o' drug design. All the recent reissues with this label are pussified attempts at being into "Drug Design".

You all are too late to get any street cred for Rational Design and you all are a bunch of lightweight poseur-Ho-Dads!

Computational guys, other than Schrodinger, should apply for positions at Home Depot and save the rest of us heartbeats while we make real drugs, using real chemistry.

Ye Gods!

Permalink to Comment

20. Anonymous on February 21, 2012 11:19 PM writes...

The article is an ACS AuthorChoice, meaning it should be free access to everyone. Derek added a link in case people cannot access the article, but that should not be the case for anyone unless they do not have an internet connection (in which case they would not be reading this blog). While the paper certainly has a computational perspective (would one expect anything else from a perspective paper by computational chemists), there are many examples cited where computational tools did not drive the selectivity design. As some people have already stated here, "rational" by definition means doing something based on reason or logic. That might involve computational chemistry, but often just requires looking at a ligand-protein structure or even analyzing SAR from a ligand series. However, with selectivity there tends to be a lot less SAR of ligand series binding to multiple targets, making the rational design part more difficult. Ideally, everyone in a project would input knowledge and insights based on their expertise and the team would proceed rationally based on all of the available information. The alternatives to proceeding rationally are not very attractive.

Permalink to Comment

21. H2L on February 21, 2012 11:32 PM writes...

#18 OldLabRat, many PK/PD and ADME/tox issues can be designed away with the right understanding of selectivity (of course, that assumes that the biological pathways of interest are known). Problems from things like hERG, P450s, and others can be avoided if the potency to those off-targets is reduced while increasing or maintaining potency to the target of interest. Not that the biology is solved or easy, but selectivity is the key to many parts of drug design beyond what we often think about when we try to tune binding to a single target of interest and away from a few very close members of the same target family.

Permalink to Comment

22. odj on February 22, 2012 12:54 AM writes...

If the Rational Drug Designers and the Evidence Based Medicine people ever get together, we will have an unstoppable force!

Permalink to Comment

23. milkshaken on February 22, 2012 2:36 AM writes...

@22: The unstoppable force - But we did already. Ever heard of Sasha Shulgin?

Permalink to Comment

24. Anonymous on February 22, 2012 3:34 AM writes...

This discussion breaks down to simply one thing: Perfect Marketing! As anyone here agrees, 'rational' is a lot more than the Scrodinger posse claims. The computational tools that are most efficient in supporting 'rational' drug design are as simple as
1.) proper 2D depictions of molecules / SAR series
2.) similarity searches of billions of molecules 3.) and the like.
Keep in mind that all these tools can help sientists but they won't supersede them at all!
Who gave those Schrodinger sales experts the opportunity to spam scientific literature once again with intriquing claims that are neither proven nor disproven? Also, it is not a listing of success stories, as no one can really measure the impact of one specific (computational) tool on success. Derek, as you said: "What would have happened if we had not used some specific tool?"

Permalink to Comment

25. NoMiracles on February 22, 2012 6:04 AM writes...

In response to entropyGain, I have been around for the same time, and I would argue that only recently has the computational world started to accept that we are data limited. Much of that is what Derek describes. Algorithmic innovation around noise is not very productive, even if you get a method paper published! FoldIt is interesting, but not all that useful. Diversity is boring, but can help with efficiency and therefore can be useful. There have been many attempts at innovation but they often fall down when applied to new problem with unappreciated variation. I think one of the best examples of this is the work that has shown just how bad crystal structures can be.
Being locking into your ivory tower of 'innovation' is sometimes best left to the marketing folks at Schrodinger and others

Permalink to Comment

26. odj on February 22, 2012 8:02 AM writes...

@23 ...unstoppable at forcing Payors to pony up. While the original post was intended to be cynical (verging on trollish), I admit that Sasha succeeded brilliantly, although not with the "Payors" that most pharmas have in mind.

Permalink to Comment

27. Mo on February 22, 2012 8:32 AM writes...

@19. MoMo
I dispute your claim. Clearly Paul Ehrlich is the father o' drug design.

In my view computational drug design isn't worth much without coupling to an experimental kd or IC50 and those aren't worth much without a coupling to EC50 and so on. You need your group of positives and your group of closely related negatives and then advance.

Permalink to Comment

28. Terry malloy on February 22, 2012 12:19 PM writes...

In the course of drug development one of the easiest things for a medicinal chemist to do is to identify potent compounds. After that, if we are not careful, we spend our time making compounds to help a computational chemist rationalize/refine their models. Can computational types design a way out of a TDI, improve PK or eliminate an unwanted transporter SAR. In end there is so much more to what medicinal chemists do beyond simply making potent compounds. IMHO it is this current mindset of potency in in vitro biochemical assays that is driving the misguided notion that one of the most complex of human endeavors - drug discovery - can be reduced to two computed values - molecular weight and cLogP. Medicinal chemistry is a craft which requires years of experience to develop. Computational chemistry has it's place and value but in te end it is a discipline that contributes to drug discovery, not one that drives it. The computational acolytes should bear this in mind as they promote their craft because if they do not it will serve to further reduce their credibility with people that, over time and with experience, have learned what can and what can not be done.

Permalink to Comment

29. cynical1 on February 22, 2012 12:35 PM writes...

Is patent busting considered 'rational design'? I've seen a lot of manuscripts that claim it is anyway.

Permalink to Comment

30. H2L on February 22, 2012 1:14 PM writes...

Clearly, knowing something about how a molecule binds to a target of interest can add insights. Furthermore, knowing how it binds to undesirable targets can add further insights. On top of that, calculations can help drive decisions to improve the efficiency of a project, even if it is by suggesting things _not_ to make. I do not think anyone would claim that calculations are perfect or could ever replace the role of the med chemist, but it is hard to argue that there is no value in understanding the underlying physics behind molecular recognition. We have successfully used information from modelers in many projects. Of course, we do not follow orders from modelers, but they play an active and constructive role in the design process.

Permalink to Comment

31. Anonymous on February 22, 2012 2:36 PM writes...

I am just curious, what are the differences between computational chemists and med chemists? I mean, shouldn't they converge overtime?

Permalink to Comment

32. pilsner on February 22, 2012 8:33 PM writes...

I am a computational chemist. I have and will continue to always avoid the rational drug design phrase. The reason is that the medicinal chemistry approach of systematically modifying a molecule by one or another functionality (methyl, ethyl, butyl, etc.) is entirely rational and should not be termed otherwise, especially in the absence of other information. Computaional chemistry is an effort to utilize available information, protein structure and pharmacophore, to help prioritize synthetic strategies. I much prefer the terms structure aided drug design or computer aided computer design.

Permalink to Comment

33. drug_hunter on February 23, 2012 6:02 AM writes...

Took me a few days to get around to it, but I've actually read the article now, and IMHO it is a perfectly good REVIEW of recent examples of ways to attempt to design selectivity into molecules. Very systematic. 184 references, so far as I can tell well selected. The authors conclude:

"We hope that the principles laid out in this work and the associated examples will help make the practice of selectivity design more transparent and lead to more explicit con- sideration of how selectivity can be improved in the process of rational drug design."

I suspect if the title of the paper had been "Systematic Approaches to Improving Selectivity" or "Structure-Guided Approaches..." then we wouldn't have even had this blog entry.

Permalink to Comment

34. metaphysician on February 23, 2012 5:23 PM writes...

I'm not a biochemist, but am I the only one who is thinking "Sounds like a good idea, maybe somebody should try it some day"? *cough*

Permalink to Comment

35. Andy on February 27, 2012 3:34 AM writes...

@30 H2L: Confronted with a 'don't make this' structure from a comp chem source, I would seriously struggle to agree, ignore the structure and move on. I'd never sleep properly again.

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