« Catalyst Pharmaceuticals And Their Business Plan |
| Size Doesn't Matter. Does Anything? »
October 21, 2013
Getting van der Waals Forces Right
I'm actually going to ignore the headline on this article at Chemistry World, although coming up with it must have made someone's day. Once I'd gotten my head back up out of my hands and read the rest of the piece, it was quite interesting.
It's a summary of this paper in Nature Chemistry, which used the ingenious system shown to measure what the alkyl-chain interactions are worth in different solvents.
The team has now used a synthetic molecular balance to measure the strength of van der Waals interactions between apolar alkyl chains in more than 30 distinct organic, fluorous and aqueous solvent environments. The balance measurements show that the interaction between alkyl chains is an order of magnitude smaller than estimates of dispersion forces derived from measurements of vaporisation enthalpies and dispersion-corrected calculations. Moreover, the team found that van der Waals interactions between the alkyl chains were strongly attenuated by competitive dispersion interactions with the surrounding solvent molecules.
There are two ways to look at this, and they're not mutually exclusive. One, which the Chemistry World article takes (in a quote from lead author Scott Cockcroft), is that this could simplify computational approaches to compound interactions, because calculating van der Waals forces is a much more intensive process. If solvent interactions are just going to cancel them out, why spend the resources? And that's true, but it brings up the other question: why did we think that vdW forces were so strong in the first place? As the quote above indicates, a lot of the experimental evidence is from gas-phase measurements, and the addition of solvent molecules clearly means that those values aren't as generalizable as had been thought. But that brings up the next question: why haven't computational methods shown before now that the gas-phase experimental data could be leading things astray?
I don't know the literature of the field well enough to answer that question, but given the sorts of exchanges that were taking place back in that recent Nobel Prize post, I'll bet that there are some people out there who can. Have there been computational methods that pointed toward the experimental data? Or have some of those efforts been directed more towards just seeing if the gas-phase data could be reproduced?
As a medicinal chemist, naturally, I'm wondering how we need to be thinking about binding of molecules to the active sites of enzymes. That's certainly not a solvent-filled environment, but inside a protein, water molecules are the bridge between being in a vacuum and being in solution. It'll depend on how many you have to worry about, what their roles are interacting with the protein and the ligand, how defined they are spatially, and how much of the molecule will be exposed to solvent. These things we already knew - will these new experimental results help us to get better at it?
Update: see the comments - lead author Scott Cockcroft says that his group is looking for computational collaborators for some of these very purposes.
+ TrackBacks (0) | Category: Chemical News
POST A COMMENT
- RELATED ENTRIES
- Merck's Aftermath
- Models and Reality
- Rewriting History at the Smithsonian?
- The FDA: Too Loose, Or Appropriately Brave?
- More Magic Methyls, Please
- Totaling Up a Job Search
- Humble Enzyme Dodges Spotlight
- Unraveling An Off-Rate