1. John Novak on September 13, 2007 9:46 PM writes...

Talk to engineers, talk to software developers. You’ll get the same perspective, accompanied by much eye-rolling and waving of arms.

Oh, sweet Jesus, yes. In Big Engineering, that is definitely the case. We have elaborate structures in place to make sure we do not make a mistake that ends up killing someone. Then there's the structures in place to make sure we don't make a mistake that costs a few million dollars....

It is miraculous that innovation happens, sometimes.

2. Clark Kent on September 14, 2007 7:48 AM writes...

I like to call it the "magic methyl" effect.

3. Lars on September 14, 2007 8:04 AM writes...

I like small changes.

It's not like they're predictable...we've all seen the effects that methyl group can have on affinity when you make that ever-so-innocent move from the ortho to the meta position. Suddenly you've got a compound that's dropped 4 orders of magnitude in Ki despite your computational chemist's assurances: "There's plenty of space in that pocket!!".

But when it comes to preclinical properties, small changes often lets you get around solubility issues, metabolism, protein binding, hERG etc without opening up a whole new can of problems that you inevitably run into if you throw on a big polar extension or your favorite "big slab of fat" (a.k.a. adamantyl).

In the end, most compounds that make it really far (I'm talking Phase II/III) are actually quite pleasant to look at. Call it "minimalistic beauty". Just enough of everything you need and none of what you don't.

"Less is more" is an impotant lesson. Not just for nouveau riche basketball players on "MTV Cribs" but for medchemists too...

4. CMC guy on September 14, 2007 10:09 AM writes...

Ah yes the wonderful life of the methyl, ethyl, butyl, futile derivatives- not sure would classify lack of enhancement as mistakes because unless it’s made and tested its hard to know if could gain improvements even with wonders of 3D modeling. Also failure of a compound in one program has been lead(s) for a different projects. You are also correct does take a certain mentality to sustain career in discovery R&D where odds are against ever finding an actual marketed drug. The best researchers I know just charge ahead and do it without long discussions on why it may or may not work. Doing Process work puts one closer to potential "drug success" and still endure the twisted pathways to solutions however the fall can be harder when a candidate fails.

I had to laugh at the retrospective insight as I do recall when writing my Thesis that felt I could have done all the work over in about 4-6 months instead of the 3.5-4 years it took (providing a labmate would stop borrowing my cleaned glass from the rack).

5. SBC123 on September 14, 2007 10:15 AM writes...

The projects that I have been involved in had different stories. The compounds we put in animals were much different from the screening hits. I used to believe this is true for most modern drug discovery programs because the sensitive and irrelavent screening methods are picking hits far from being efficacious and PK friendly (traditional leads). Maybe I am wrong then. Half of my projects actually needed scaffold level modifications.

6. WC on September 14, 2007 12:52 PM writes...

I've worked on several programs that have entered clinical trails and if you look at the initial leads and compare them to the final candidates, the modifications made weren't elaborate or profound, except for one of them. At a former employer, we used to have these brainstorming meetings where we were encouraged to come up with brand new leads for existing programs. Now, I'm all for new off the wall ideas, but for the most part, I came to dread these meetings. I saw way too many ideas pursued that went nowhere. I saw much more success with sticking with initial leads and keeping it as simple as possible.

7. milkshake on September 14, 2007 10:33 PM writes...

Scaffold hopping is a worth doing if you
a) have no good lead compound - and you hope to get lucky by trying a little bit of everything
b) you have a major trouble with stability, solubility, poor selectivity or cell potency etc and you are so desperate that you think you won't be able to solve them by simple substitution changes/side-chain etc.
c)you have a good co-crystal X-ray data that gives you plenty of new ideas for dramatic improvements

The situation in point c) it treacherous because unless you have explored enough of simple modifications from the methyl-ethyl-allyl-cyclopropyl-fluoro-chloro-trifluoromethyl-methoxy category, you can end up doing big leaps of faith based on the X-ray picture, get messy and incomplete SAR and miss something important

8. Petros on September 15, 2007 2:24 AM writes...

There were some articles about the "Magic Methyl" effect quite a while ago (15-20 years)

9. bcpmoon on September 17, 2007 5:36 AM writes...

Ok, it was frustrating to realize that I could have done the work for my thesis in three weeks instead of years, if I had known at the beginning what I did at the end, but I found it always rewarding in itself that my end could be a beginning for somebody else, going forward into the unknown. I love science.

10. DrSnowboard on September 17, 2007 11:40 AM writes...

Scaffold hoping (yes, the spelling is deliberate) always reminds me of that Irish joke - the one where the guy can't figure out how to give directions so he says "Well, to be honest, if you're going to Dublin / Balbriggan, I wouldn't be starting from here.."
But then again, "Little do ye know your own blessedness; for to travel hopefully is a better thing than to arrive, and the true success is to labour."

11. xchemist on September 20, 2007 11:27 PM writes...

DrSnowboard: I understand the sentiment, but truly I would rather arrive than travel hopefully, having done both. More time then to visit with the ones one wanted to see. But the trick of it is, how can one substantially ameleiorate the problem-- efficacy and tox risk-- and still have new composition of matter to get a product? The holy grail of the med chemist; maybe it's possible, or not.

12. MikeyMedchem on October 27, 2007 4:08 PM writes...

Does anyone have any of those references Petros mentions? They'd be very helpful to me. It's amazing that much of what we med chemists do is really based on folklore such as "magic methyl"...and we follow these premises by word of mouth instead of actually seeing the research!

13. milkshake on October 28, 2007 3:15 PM writes...

Look up the potency of cis-3-methyl fentanyl vs the plain olf fentanyl, there is >20 potency improvement with the (-) cis enantiomer. It is about 1 thousand times more potent then heroin. (Fentanyl is only 50 times more potent). Racemic cis 3-methyl fentanyl was used in the Nord-Ost theatre siege in Moscow. Since no antidote was administered to the hostages, more than 100 of them died of overdose. (The OMON troops had the antidote for their own use - but were not allowed to say that Naloxone/Naltrexone would work because it was such a secret).