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

« Intermediates | Main | Live-Blogging Arena's FDA Committee Hearing »

September 16, 2010

Six Sigma in Drug Discovery? Part One - Are Chemists Too Individual?

Email This Entry

Posted by Derek

I had an interesting email about a 2009 paper in Drug Discovery Today that has some bearing on the "how much compound to submit" question, as well as several other areas. It's from a team at AstraZeneca, and covers their application of "Lean Six Sigma" to the drug discovery process. I didn't see it at the time, but The title probably made me skip over it even if I had.

I'll admit my biases up front: outside of its possible uses in sheer widget-production-line settings, I've tended to regard Six Sigma and its variants as a buzzword-driven cult. From what I've been able to see of it, it generates a huge number of meetings and exhortations from management, along with a blizzard of posters, slogans, and other detritus. On the other hand, it gives everyone responsible a feeling that they've Really Accomplished Something, which is what most of these managerial overhauls seem to deliver before - or in place of - anything concrete. There, I feel better already.

On the other hand, I am presumably a scientist, so I should be willing to be persuaded by evidence. And if sensible recommendations emerge, I probably shouldn't be so steamed up about the process used to arrive at them. So, what are the changes that the AZ team says that they made?

Well, first off is a realization that too much time was being spent early on in resynthesis. The group ended up recommending that every lead-optimization compound be submitted in at least a 30 to 35 mg batch. From my experience, that's definitely on the high side; a lot of people don't seem to produce that much. But according to the AZ people, it really does save you time in the long run.

A more controversial shift was in the way that chemistry teams work. Reflecting on the relationship between overall speed and the amount of work in progress, they came up with this:

Traditionally, chemists have worked alongside each other, each working on multiple target compounds independently from the other members in the team. Unless managed very carefully by the team leader, this model results in a large, and relatively invisible, amount of work in progress across a team of chemists. In order to reduce the lead time for each target, it was decided to introduce more cooperative team working, combined with actively restricting the work in progress. The key driver to achieve and sustain these two goals was the introduction of a visual planning system that enables control of work in progress and also facil-
itates work sharing across the team. Such a visual planning system also allows the team to keep track of ideas, arrival of starting materials, ongoing synthesis and compounds being purified. It also makes problems more readily recognizable when they do occur.

We have reflected on why chemistry teams have always been organized in such an individual-based way. We believe that a major factor lies in the education and training of chemists at universities, in particular at the doctoral and postdoctoral level, which is always focused on delivery of separate pieces of work by the students. This habit has then been maintained in the pharmaceutical industry even though team working, with chemists supporting each other in the delivery of compounds, would be beneficial and reduce synthesis lead times.

OK, that by itself is enough to run a big discussion here, so I think I'll split off the rest of the AZ ideas into another post or two. So, what do you think? Is the "You do your compounds and I'll do mine" style hurting productivity in drug research? Is the switch to something else desirable, or even possible? And if it is, has AstraZeneca really accomplished it, or do they just say that they have? (Nothing personal intended there - it's just that I've seen a lot of "Now we do everything differently!" presentations over the years. . .) After all, this paper is over a year old now, and presumably covers things that happened well before that. Is this how things really work at AZ? Let the discussion commence!

Comments (50) + TrackBacks (0) | Category: Drug Development | Life in the Drug Labs | Who Discovers and Why


1. NYSpursFan on September 16, 2010 9:23 AM writes...

Exceedingly interesting hyppothesis. It's certainly true in any PhD program that you're trained to think and act and produce and write as an individual. Then you get a job in industry and are thrown into a "team" environment, which isn't really a team, per se, but a group of individuals. The group is being managed by an "individualist" with no experience or training in group oversight or management. As I think out loud here, there may be quite a bit of truth in this. I also see this in the financial and consulting communities. Perhaps the answer lies in "team" PhDs?

Permalink to Comment

2. Bob on September 16, 2010 9:36 AM writes...

If this was true, then surely "german model" organisations (Novartis, Merck-Serono, Roche et al.) would have gone out of business years ago or have been demonstrably less productive than the competition.

In addition to an over-abundance of six sigma black belt crapola, AZs real problem is evidently (and by their own admission) a lack of careful med chem managers.

Permalink to Comment

3. Cellbio on September 16, 2010 9:43 AM writes...

Nothing here specific to Lean or Six Sigma. I worked in a great team culture in my first company, and had to learn that my productivity was in the team and not my lab. The culture was built by the founders, long before six sigma became popular, and without meetings and with no requirement to laminate anything. it was all supported by the way the leadership acted.

Also, nothing particular to chemists way of working. In fact, I found the biologist, or target champion, to be much worse than the chemists. Each wanted to screen their teams compounds at the dose they chose, against the related targets they thought reasonable and in assays they built, even if closely replicating an existing assay. This meant chemists compounds were decorated with non-overlapping data sets from similar but importantly different assays that made cross-team comparisons difficult.

I fixed all this with common sense, and occasionally a bit of insistence, but never had to invoke Lean or SiX sigma. In fact, my best allies for change were the chemists, who wanted a better system of assuring rapid and meaningful turn-around of data that informed the next round of synthesis. I found that most every chemist had the experience of working in a team with assay 'blinders' which resulted in months to years of work that would have been redirected if only a simple assay had been run early in the program. Easy fix with sound rationale, no need to bring in the black belts.

Permalink to Comment

4. Hap on September 16, 2010 9:44 AM writes...

1)...and credited as an individual. From grad school, part of what potential employers want to know is what the capacities of each person are, and an education that makes it hard for employers to disentangle the contributions of each person will probably make it (more) difficult for those people to get jobs afterwards.

2) I don't know the day-to-day operations in drug research, but I had assumed that people mostly made separate and discrete things, or interacted on separate steps in a synthesis (though even that means that unless you have lots of things going at once, one person's difficulties hinder everyone else in the sequence). Even in a team environment, people do discrete tasks, and anything that lumps lots of things together makes their output dependent on multiple people in concert and makes the contributions of individuals difficult to disentangle. How do you change this?

Permalink to Comment

5. Calvin on September 16, 2010 9:48 AM writes...

Hmmmm. I've seen the AZ Six Sigma folks in action. It's a cult for sure. It's a career badge thing on their journey up that career ladder. However, I'd agree that the combative nature of many PhD groups does not lend itself well to teamwork and there is scope to do things better (Chris Lipinski used to talk about this too). In fact, we use the tracking board idea but not nearly as rigidly as AZ and it does work pretty well. Certainly, it's helped the group that I work in share information and intermediates more readily to such an extent that specific chemists will do a step, pass it on, and then maybe get the batch back later on. Everybody can see where everybody is (roughly) with a target or intermediate. So it definitely helps but only because it's a simple tool that aids communication. I have to admit I was VERY skeptical and we certainly don't use it as AZ do (we don't date when a compound goes into synthesis and when it gets submitted and log the overall time) but it's something that's worth trying out

Permalink to Comment

6. SixSigmaSkeptic on September 16, 2010 9:50 AM writes...

Hi Derek,
Having worked at GSK for some years where some groups were very keen on SixSigma etc, I think you've summed it up perfectly in your second paragraph. SixSigma was the perfect vehicle for ambitious but scientifically/technically limited managers to climb the career ladder. It was a Godsend for them, lots of pointless meetings, full of meaningless buzzwords that produced nothing tangible except promotions for said managers. Me personally, I wouldn't hire anyone with SixSigma tendencies, for the same reasons, I'm also very suspicious of such titles as "project managers" and "business analysts", "IT strategy managers" etc. These types of roles appear to have proliferated in "Big Pharma" in recent years and are I believe major factors in its current decline! That's my very strong personal bias born of witnessing the rise and rise of these "empty suits"!

Permalink to Comment

7. Cloud on September 16, 2010 10:38 AM writes...

Like @cellbio, my perspective is more on the overall industry than just on chemists. I think the switch from focusing purely on your own project to thinking more broadly about the good of the organization is one of the hardest things about the academia to industry transition for some scientists. I'm in biotech, where projects are often competing for limited resources. I've seen too many project leads who are so focused on the success of their projects that they kill/maim other important projects. I just want to shake the project lead and say "Yes, your project will succeed. But what good is that if the entire company is out of business?"

Permalink to Comment

8. dearieme on September 16, 2010 10:49 AM writes...

Those engineers who do PhDs also do them as individuals. Yet afterwards they tend to work in well-integrated teams.

Permalink to Comment

9. LeeH on September 16, 2010 11:04 AM writes...

As a cheminformatics guy I'm usually in favor of slick ways to organize information and make things easier for project teams. That being said, my experience is that you can't get people to use any system unless you can demonstrate that the tool will make life easier for them, and not just add another accounting burden. My sense is that if people at AZ (or elsewhere) are actually using this kind of system then it is really making their life easier in some tangible way (or they've been threatened with consequences if the system isn't used - also possible).

You could imagine of how a formal system to show people where they are in terms of the synthesis of compounds with respect to others in the project if:

1. They're physically separated. (Most likely)
2. They don't like to talk amongst themselves but don't mind interacting with a computer system or bulletin board. (i.e. the company tends to hire bright but otherwise socially inept)
3. The manager is very hands off and the troops have decided to organize on their own. (not so unlikely given the lack of management skill by scientists in general)
4. Combinations of the above.

Permalink to Comment

10. Petros on September 16, 2010 11:22 AM writes...

I've seen them present the virtues of this at 2 or 3 med chem meetings and expound how it had been widely adopted throughout AZ

Of course AZ is now much leaner than before having shuttered Charnwoood, Lund and various other sites

Permalink to Comment

11. cynical1 on September 16, 2010 12:07 PM writes...

The only way to encourage true team efforts in drug discovery is through team rewards. After 25 years in the industry, I can say that it just doesn't happen, particularly in larger organizations. Individuals are rewarded even when there was a true team effort that produced the accomplishments. Typically the project leader receives a disproportionate amount of the accolades even if their main contribution was making nice Power Point presentations for upper management or worse, an active impediment to progress. You don't have to have an MBA or a Lean Sigma Six black belt to understand that you have to incentivize teamwork. "Team pride", as they put it, doesn't pay the bills or bring job security and while I don't work at AZ, I'll bet that this little exercise has done little to address those two critical issues.

One other observation: there are 25 authors from AZ on that paper. That's just so wrong on so many levels.

Permalink to Comment

12. victory pilsner on September 16, 2010 12:16 PM writes...

It may be fine for development and manufacturing although I doubt it. Six sigma stuff is garbage for research and the epitome of what's wrong with big pharma. It is a lowest common denominator approach that is more bureaucracy, chaff, and stifling of creativity - no wonder innovation comes predominantly out of small business. I'm with Dilbert -

Permalink to Comment

13. DLIB on September 16, 2010 12:42 PM writes...

Sounds alot like the work in process issues in a Semiconductor Fab being translated into the chem lab.

Permalink to Comment

14. processchemist on September 16, 2010 12:47 PM writes...


if for "development" you mean chemical development, six sigma is deleterious. In the past I've seen more than a chemical development project managed with "productivity improvement" strategies, with the poorest results. You can use parallel systems, but parallel/automated chemistry increases productictivy only if there's a chemist who really knows what he is doing (single parameter variations, fancy parallel systems with non scalable results can be a total waste of time). Same for DOE: cool software can be a solid waste of time, or a wonderful way to demonstrate the obvious, if stuffed with meaningless parameters or used for the search of third order effects that approximate zero.
And I've seen ALL of this kind of useless combinations (great - often hyped - tools and poor process chemistry).

Permalink to Comment

15. Khaki Mango on September 16, 2010 1:00 PM writes...

Lean process improvements work very well for repetitive, infrequently-changing, and quantifiable processes. I've seen it work very well in compound management and/or HTS screening environments. It can also work well in specific areas of the lead optimization process that are done routinely, as noted above, like in compound submission, standardized assays, etc; however, by the nature of the beast, the discovery/lead optimization process has too many variables that must change many times in order to be successful including assay development on the fly, changing SAR/focus on the chemistry side, changes to the screening trees, etc. Some of the best work I've seen and done has been outside of the "standard" process or thinking. Creativity and six sigma just don't go together; we're all too different.

With regards to compound submission amounts, years ago I mandated 30 mg as the target amount (a goal, not a hard rule) for lead op work to the whole group, and was met with complaints and grumbles; but, it greatly paid off on the back in by always having enough to do all in vitro work and pk/other in vivo expts. Of course, then there was trying to actually get a pk slot or animal study in a timely manner.

Permalink to Comment

16. Dave L on September 16, 2010 1:18 PM writes...

I'm a PhD Chemist and was very skeptical of SixSigma when I was "forced" to take the 3 week course several years ago. I worked in a manufacturing area that was adopting the SixSigma methodology and we were all encouraged to take the training. I have to say that I've successfully used the SixSigma process a few times in my career. If you know how to use the tools, are good at them, and are not distracted by management wanting to rush it along, the process can really help in bringing a wide group of people together with very different skills and abilities and maintain their focus on solving a problem. I do find it better suited for manufacturing problems such as decreasing defects in a process. I haven't seen any great examples of enabling innovation and those sorts of things. The biggest problem is half-hearted adoption of SixSigma by managment only wanting to look like they are "on board". They roll out the training, establish some projects, and then quickly claim accomplishement with liberal use of "buzz words" to demonstrate they are "on board". But there's no real commitment. Few things succeed with no commitment.

Permalink to Comment

17. Skeptics on September 16, 2010 1:28 PM writes...

Re: Are Chemists Too Individual?

Ha! Its obvious none of them can think for themselves at all.

Definition of a Med Chem: Design of small-molecule drugs to target proteins and inhibit their biological function.

Reality: Not a single med chem has an expert understanding of protein-ligand associations. Therefore target specificity and efficacy is never improved upon. This simple fact does not appear to bother a single med chem.

There is no bedrock to even call this a profession. So it is this investors opinion that the med chems and the six sigma kooks deserve one another.

Permalink to Comment

18. Mutatis Mutandis on September 16, 2010 1:58 PM writes...

The problem with this type of optimization strategy is that it tends to split up tasks in sub-tasks and sub-sub-tasks, so that each little job can be executed by a dedicated specialist who presumably excels at it; rather than the whole being accomplished by an all-round scientist who is supposedly "good enough" at any of the sub-sub-tasks.

That often sounds more efficient on paper, but the reasoning often ignores the communication overhead, the loss of motivation and creativity if people don't have a feeling of ownership of the project, and last but not least the difficulties it creates in cooperating across disciplines. At some level chemists, cheminformaticians, assay biologists, and people with clinical and commercial insights will have to sit together. If these are all going to be teams rather than individuals, the sum of them is going to look like a parliament instead of a drug discovery team.

Nevertheless, up to a given level, it can work. It is just important to keep an eye on the cost/benefit curve, and not overdo it. I think that e.g. if you have 20 chemists, it is better to make 3 to 5 teams out of them, rather than one big team.

As for six sigma, it has its value and its downsides. The real problem may be that most people in pharmaceutical R&D know shockingly little about statistics, and therefore tend towards uncritical acceptance of such tools.

Permalink to Comment

19. Kay on September 16, 2010 2:16 PM writes...

I switched from a large pharma company, where everything was done individually, to a small biotech where synthesis was done as a group. There were good and bad points to both approaches. The problem with working on a "group" project is that the person who makes the final compound often gets the credit for the work, since their name is on the compound. If you are a lower level chemist, you may get shuttled off to do scale-up, and if you stay there long enough, it can derail your career. You may be doing work that is crucial for the team, but it tends to be invisible work that may not be recognized when it comes to raises and promotions. At the small company where I worked, it led to a lot of competition and infighting among the chemists.

Permalink to Comment

20. Anonymous on September 16, 2010 2:28 PM writes...


You'll probably get flamed enough by the rest of the med chemists, so I'll avoid the Molotov cocktails.

Though you state an inflammatory generalization, there is some truth in the fact that we do not understand protein-ligand interactions at the expert level. That is because only now are very crude tools to model those become available with the computational speed to assess even small molecule interactions statically, much less dynamically.

It is helping the field to have much more structure-based design going on and it is becoming a good tool in the toolbox (cliche alert!) to use. However, it is still crude, has its limitations, and many of us have seen it sometimes work but often not, so we still use traditional empirical SAR techniques and experience to drive the programs.

Reality: Target specificity and efficacy are constantly being improved upon at least around here and throughout the industry, so I don't know where you're getting your info from. We're just not as good at this drug thing as we wish we were.

Permalink to Comment

21. Jack Vinson on September 16, 2010 2:28 PM writes...

I don't know directly about AZ and this particular implementation, but if you remove the buzzwords that get in people's way, the paragraph you clipped is a reasonable approach.

Most organizations have far too many things going on at one time. Businesses and managers have no gating process: if they have a good idea, they just go ahead and start it under the assumption that the sooner you get started, the sooner it will finish. The problem is that with so many projects, there are just too many moving pieces, and no one can determine what is the next right thing to move the projects forward. Or it becomes too easy to jump to the next interesting thing, rather than the next thing that needs to happen.

Permalink to Comment

22. CYTIRPS on September 16, 2010 2:37 PM writes...

Six Sigma is just another excuse used by the R&D air heads to pretend that they are doing something. Six Sigma is only a QC strategy for manufactoring a product. We do not know which compound is the final product in discovery research! If the project managers can predict the final product, they can skip research and go to manufacturing directly. No one is that good yet. It is a good thing for individual chemist to explore in different direction. There is no road map leading to the final product. The important thing is to look at the data at the end of the day and determine whether it is the right direction. A common mistake is that many chemists attach to their own ideas and ignore the data. IMHO, the only useful tool in Six Sigma for discovery is the PICK chart. However, even that is just common sense.

Permalink to Comment

23. partial agonist on September 16, 2010 3:35 PM writes...


"Not a single med chem has an expert understanding of protein-ligand associations."

That hinges on your interpretation of the word "expert". I think we certainly have a far better grasp on fundamental concepts like productive ligand-protein van der Waals interactions, H-bonding, dipole-dipole interactions, pi-pi,cation-pi interactions, implications of ligand desolvation penalties, etc than we ever did before, and can often apply such concepts in a rational way to improve ligands.

"Therefore target specificity and efficacy is never improved upon."

Well, the biologists told us forever that kinases, for example, were poor drug targets because we could never get selective within the ATP-binding pocket. Do you want to see the data for a kinase inhibitor that has subnanomolar affinity for the two kinases it was designed to hit and >1 uM afinity for 300 other kinases looked at, and also no activity vs. 100 other GPCRs and enzymes? Chemists can make amazingly target-selective compounds. The issue, I admit, is knowing what are the off-targets to avoid, and curiously what off targets actually help you when you take our precious molecules in vivo. If anything, molecules often get ruined due to too much target selectivity, so in vivo a single mutation wipes away all efficacy. We all know the "dirty" lead that works in vivo, then we polish it up against the "target" and improve potency & selectivity, then it goes back to animals and fails. Well... duh... the initial "dirty" lead was hitting multiple targets of value.

"This simple fact does not appear to bother a single med chem."

Just a load of crap: Target selectivity is something we think about every day and in every structure we make. We stay up night thinking whether the structural changes I want to make will affect CYP profiles, hERG channel, half-life, or 100 other things.

"There is no bedrock to even call this a profession"

How are those gene patents and pathway patents working out? What matters is the molecule you can put in a bottle. You need chemists for that, yesterday, today, and as far as I can tell, always.

Permalink to Comment

24. Chris on September 16, 2010 4:41 PM writes...

My feeling is that basic research changes too frequently for it to be useful. On one project we had to make a candidate quickly to access an unexpectedly available safety slot, we had the whole team working on the route using a flow chart to monitor progress on multiple parallel strategies. Chemists switched from one route to another to help address problems.
On many occasions we had individuals making large amounts of intermediates that were used by the whole group.
I didn't think of it as anything unusual, perhaps we needed to think of a catchy title?

Permalink to Comment

25. cliffintokyo on September 16, 2010 8:27 PM writes...

Nice balancing act, Calvin.
"Communication" says it all, for me anyway.

Permalink to Comment

26. Don Corleone on September 16, 2010 9:04 PM writes...

In the latest issue of Fortune magazine, there is an article about 3M. It talks about how innovative culture at 3M was negatively impacted during the Jim McNerney era, an MBA (one of GE "wonder" boys). One of scientists in the article said something which is so true "you can't schedule innovation". Drug discovery and six sigma do not mix.

Permalink to Comment

27. Peter Ellis on September 17, 2010 1:58 AM writes...

"chemists supporting each other in the delivery of compounds"

I'm no chemist, but this feels to me like putting nine women in a room and telling them to produce a baby in one month.

Permalink to Comment

28. SandwichMan on September 17, 2010 2:51 AM writes...

I largely agree w