[phenixbb] Is it possible to do Molecular Replacement for synthesized "small" molecule X-ray data?

Anastassis Perrakis a.perrakis at nki.nl
Tue Dec 8 12:41:34 PST 2009

What a nice molecule according to http://en.wikipedia.org/wiki/Cucurbituril 

Did you by any chance crystallize it as a salt? In the previous century,
this happened to me with a cyclodextrin complex that had a Br atom.
The Br was visible in the native patterson, and that had at the time
triggered my interest in 'ARP' that was soon after to become 'ARP/wARP'.

If there is a chance for a 'heavy' atom there, use it!

Looking at the molecule I would expect a cloud of solutions in MR, all  
pretty good but every ~50 degrees
on gamma, given the cyclic molecule, if I got the structure right. So,  
you would have no 'unique' solution.
However, putting any of the top solutions in 'ARP/wARP map improvement  
by free atoms addition and deletion' should/could yield
a good map with 1.5 A data.

Apologies for the non-phenix shameless self-promoting plugin, but I  
think you might find this actually useful.


On 8 Dec 2009, at 21:18, Randy J. Read wrote:

> Dear Yves,
> That's probably getting close to the limit of the ratio of the size  
> of the
> molecule to the resolution, which is one of the main things  
> determining
> whether molecular replacement will work. At 1.5A, you'd have to have a
> pretty accurate model of the structure as well (i.e. RMSD signficantly
> under 1A, at a guess). Another problem with molecular replacement  
> for small
> molecules is that small molecule crystals tend to be tightly packed,  
> which
> makes things somewhat harder. Nonetheless, I've done test  
> calculations in
> the past that show it can work.
> One thing you should watch out for: Phaser only accepts atoms that are
> given as ATOM records in a PDB file, so if you have HETATM records,  
> those
> atoms will be rejected.
> If you'd like to send the logfile from Phaser (preferably off-line), I
> could take a look at it.
> Regards,
> Randy Read
> On Dec 8 2009, Yves Wang wrote:
>> Hi PHENIX users,
>> We've got a synthesized molecule, called Cucurbit[7]uril, molecular
>> weight 1163 Da. We were able to grow crystals of cucurbit[7]uril with
>> one of its ligand. This crystal diffracted to 1.5A at our X-ray  
>> machine.
>> The dilemma is that, at this resolution, we can't use small molecule
>> crystallography softwares to solve the structure (correct me if I'm
>> wrong, we need at least 1A resolution?), while macromolecular X-ray
>> softwares are really not designed for synthesized molecules.
>> We have collected diffraction data of this crystal with decent  
>> quality.
>> So I'm  wondering if there's any way we can put a known crystal
>> structure of cucurbit[7]uril and solve the structure of the  
>> complex. We
>> have tried AutoMR but the solution is not very encouraging.
>> Any suggestion will be gratefully appreciated.
>> Best,
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb

More information about the phenixbb mailing list