Corante

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: derekb.lowe@gmail.com Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Emolecules
ChemSpider
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
PubChem
Not Voodoo
DailyMed
Druglib
Clinicaltrials.gov

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
Kilomentor
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
ChemBark
Realizations in Biostatistics
Chemjobber
Pharmalot
ChemSpider Blog
Pharmagossip
Med-Chemist
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
SimBioSys
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Business|Bytes|Genes|Molecules
Eye on FDA
Chemical Forums
Depth-First
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
FuturePundit
Aetiology
Gene Expression (I)
Gene Expression (II)
Sciencebase
Pharyngula
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net


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


Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem


Politics / Current Events
Virginia Postrel
Instapundit
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

« Friendly, Welcoming Triazides. No, Really. | Main | Sanofi's Feeling Just Wonderful »

June 6, 2014

The Robustness Test: Why Don't We Just Require It?

Email This Entry

Posted by Derek

There's a follow-up paper to that one on the robustness of new synthetic methods that I blogged about here. This one's in Nature Protocols, so it's a detailed look at just how you'd go about their procedure for shaking down a new reaction.

The reaction they've chosen is a rhodium-catalyzed indole formation (phenylhydrazine plus substituted alkyne), which is a good test bed for the real world (heterocycles, metal-catalyzed mechanism). The authors suggest a matrix of additives and reaction conditions, analyzed by GC, as in their original paper, to profile what can be tolerated and what can't. It's good to have the detailed reference out there, and I hope it gives referees and journal editors something to point at.

But will they? I can imagine a world where new reactions all have a "standard additives" grid somewhere in the paper, showing how the yields change. You could even color-code them (the old stoplight slide scheme, red/yellow/green, would be fine), and then we'd all have a way to compare synthetic methods immediately. Aldrich and others could sell the pre-assembled kit of the standard compounds to use. This would also point out reactions where more useful work could be done, since it would be immediately obvious that the new Whatsit Cyclization fails in the presence of tertiary amines, etc. Too often now you have to work that our for yourself, usually by seeing what the authors left out.

So why don't we all do this? It's more work, that's for sure, but not an incredible amount more work. If the major synthetic chemistry journals starting asking for it, that would be that. It would also make the publication landscape even more clear, because the titels that don't ask for an extra few days to be spent on the reaction conditions would be hard-pressed to make a case that they weren't just venues for hackwork (or for people with something to hide). I'd rather read about reactions with a clear statement of what they'll work on and what they won't - wouldn't you?

Comments (13) + TrackBacks (0) | Category: Chemical News | The Scientific Literature


COMMENTS

1. John on June 6, 2014 8:59 AM writes...

I support this idea, but i can see there is a few potential problems as well. And it is partially overlap with the substrate scope as well.

Permalink to Comment

2. Phil the Aldrich Guy on June 6, 2014 9:35 AM writes...

In order for there to be a general kit, there needs to be agreement on the full spectrum of functionality that should be tested. After all, one reaction's additive is another reaction's substrate. This becomes problematic for a commercial kit. There will always be one or two "additives" that a customer will look at and say they don't need.

Permalink to Comment

3. CMCguy on June 6, 2014 1:42 PM writes...

I remain skeptical of both ability to influence such expanded reaction profiling as a conventional practice and if would actually be of significant value. Its not only synthetic journals that would need to require but more critically funding agencies and prominent leaders in the field that would have to support and produce routinely. Lab work requires manpower, and possibly could outsource to undergrads or first years or even robotic systems in cases (per combichem days), yet IMO not likely the most effective use of time and funds for training more mature experienced researchers.

I also worry in chemistry that "not an incredible amount more work" can turn frequently out to be black hole even if when most reactions appear to work (providing you can even trust only GC to provide realistic yield or confirm expected product is made without undergoing other chemistry which increases with more functionality) as can never do everything that can come up with on paper and how hard can you work on failures to find conditions more suitable. Substrate choices get limited by availability and cost factors with high probability that still will not adequately probe. More often than not even good collections like Organic Reactions rarely have examples that are the exact enough pieces so act as hints at what might or might not work until you try it to demonstrate works in your reaction. Will standard additives become mandatory so people focus on those and not think outside the box?

While I do appreciate the concept of expansion of reported variability and known restrictions in initial publications, I would really favor doing more research on introducing novel agents to solve a different problem (or same one differently), especially as a process chemist who has had to apply literature procedures without solid confidence would fit.

Permalink to Comment

4. Ted on June 6, 2014 2:08 PM writes...

Hi:

GC yields?

2 steps forward, 1 step back.

-t


Permalink to Comment

5. Dave on June 7, 2014 8:18 PM writes...

Why all the hate for in situ yields analyzed by GC (or NMR, or HPLC, etc.)? How are these unreliable? I would much prefer a GC yield with multiple repeats and a rigorous calibration curve than another "isolated yield" of It all depends on what you're trying to accomplish with your research. If you're developing a catalytic reaction, and you want to know how effective the catalyst is (or what the other reaction products are), analyzing the crude reaction is way better. Workup and purification are involved steps, and unless you're doing something on a reasonable scale, I would never use an isolated yield as a measure of catalyst efficiency or chemoselectivity. Not to mention that actual mechanistic insight can ONLY be gleaned by looking at time-dependent concentration profiles, which (guess what!) are measured by things like GC/NMR/HPLC/UV-Vis etc.
The whole idea that isolated yield is the only thing that matters is horseshit. Sure, if you're just trying to make a gram or ten of material, or scaling up further, isolated yield is king. But if you're actually trying to LEARN something about the chemistry (which is supposed to be what academic research is about), in situ yield determination can give you way more information in a much shorter time frame.

Permalink to Comment

6. Dave on June 7, 2014 8:24 PM writes...

addendum to above, it should read:
"I would much prefer a GC yield with multiple repeats and a rigorous calibration curve than another "isolated yield" of less than 10mg of material."

Permalink to Comment

7. Chemist on June 8, 2014 10:11 AM writes...

I like the concept of a robustness screen, but I also worry it could prevent further exploration in complex systems. A robustness screen of an RCM reaction with Grubb's catalyst would indicate that basic nitrogens are not compatible if, say TEA was added. But certain substrates with basic nitrogens (especially sterically hindered ones) do in fact work very well in RCM reactions. As with any bit of data, people may or may not read to much into it. Although, I still do think this type of analysis is overall useful (and perhaps could identify new reactions if an additive alters the reaction pathway).

Permalink to Comment

8. sepisp on June 9, 2014 3:17 AM writes...

Har har. The whole scientific publishing system has been shifted to favoring bold, but equally vacuous claims and shameless self-promotion, when impact factors and other metrics became more important than the search for knowledge. A scientists no longer thinks "how can I improve my own understanding of nature, and share it best to fellow scientists to deservedly earn their respect". No, it's "how can I get impressive-sounding results (regardless of whether they objectively are) and get all papers published in Nature to get promotions". Actual robustness tests would obviously curtail making bold claims about near-universal applicability. This is why the scientist wouldn't want to send their article to a journal that requires it, and that's why journals are not very motivated to start requiring it.

Arcane conventions that #5 points out and which substitute accuracy theater for accuracy are on top of this.

Permalink to Comment

9. CMCguy on June 9, 2014 12:11 PM writes...

#5 Dave although I agree context and projected usage matter greatly in how to determine yield, and there is indeed ignored issues associated with giving a single isolated yield as the perceived standard (especially in mg ranges), you ask "How are (GC/Instrument yields) these unreliable?" and believe hint at answer because how often do reports involve "multiple repeats and a rigorous calibration curve" in early new technique/reagent introduction papers? Such techniques are powerful when working with known are well characterized compounds (and purity) but to be reliable there is a significant amount of upfront analytical development required, even to demonstrate the main peak is not hiding many impurities. It may be a different kind of faith in results but at least with isolated materials more often there is additional characterization to confirm structure and effectively assess purity. If a citation only provides GC yields I typically feel more work necessary to apply(which I am willing to do when circumstances require for success).

Permalink to Comment

10. Doug Steinman on June 9, 2014 12:52 PM writes...

It perhaps would be more interesting and more useful if a way was set up to post results to a common site (something like a Wiki page maybe) that one obtained when trying the new reaction in their own work. This might run into IP issues so maybe the reaction could be described as being done with "an aniline with an orhto chloro group" instead of posting an actual structure. The authors of the original paper as well as all other interested parties could then see the scope of the reaction as experienced in actual practice. I have no idea how difficult this would be to establish but I think it would represent a valuable addition to the synthetic literature.

Permalink to Comment

11. Doug Steinman on June 9, 2014 12:54 PM writes...

It perhaps would be more interesting and more useful if a way was set up to post results to a common site (something like a Wiki page maybe) that one obtained when trying the new reaction in their own work. This might run into IP issues so maybe the reaction could be described as being done with "an aniline with an orhto chloro group" instead of posting an actual structure. The authors of the original paper as well as all other interested parties could then see the scope of the reaction as experienced in actual practice. I have no idea how difficult this would be to establish but I think it would represent a valuable addition to the synthetic literature.

Permalink to Comment

12. Hap on June 10, 2014 11:38 AM writes...

Ultimately, I think most people want new synthetic methods to help them make something. Comparison of a new method with previous methods is helpful, but there are probably enough sources of variability in substrates that your mileage is probably going to vary anyway. If you can't purify the product from the rest of what's in the flask (unless you can carry it on crude without penalty, which is not universal), or lose a lot of it in doing so, then the method is not going to be better than other methods, even if the analytical data says so.

One previous example of such problems was in Hudlicky's Art of Synthesis, where the syntheses of pyrrolizine alkaloids often cited high yields without providing evidence of isolation, probably because they couldn't (the alkaloids oxidize rapidly and so are hard to handle). The cited yields were thus not reliable to anyone else; comparing yields of product to those of oxidized gorp is not a useful comparison.

I wonder sometimes if there is not too much emphasis on new chemical methods and not enough on actually purifying compounds with reduced losses.

Permalink to Comment

13. cliffintokyo on June 11, 2014 8:34 AM writes...

Perhaps we should create an enhanced version of Organic Syntheses, as a guide to things that really work. How about sub-sections for heterocyclic syntheses and metal-catalysed reactions? Others?

Permalink to Comment

POST A COMMENT




Remember Me?



EMAIL THIS ENTRY TO A FRIEND

Email this entry to:

Your email address:

Message (optional):




RELATED ENTRIES
How Not to Do It: NMR Magnets
Allergan Escapes Valeant
Vytorin Actually Works
Fatalities at DuPont
The New York TImes on Drug Discovery
How Are Things at Princeton?
Phage-Derived Catalysts
Our Most Snorted-At Papers This Month. . .