[phenixbb] Asymmetric unit content

Roger Rowlett rrowlett at colgate.edu
Mon Dec 12 07:28:13 PST 2016

The Matthews coefficient is relatively efficient at distinguishing 
likely ASU content for N<=4.  However, the probability distribution of 
the solvent content of protein crystals is broad enough that it is 
difficult to distinguish from the Matthews coefficient alone if an ASU 
has say, 4 or 6 (or 6 or 8) molecules in the ASU. Realistically, the 
Matthews coefficient is going to give you a range of likely 
possibilities. 1 vs. 2 is easy to distinguish. Not so 4 vs. 5 vs. 6.

Ultimately, you have to look at your MR solution and the molecule 
packing in the unit cell. Said packing should make chemical sense, and 
reveal highly symmetric solvent channels. A periodic "gap" in your 
crystal packing, supported by un-filled electron density, is a clue that 
something might be missing. At 7A resolution, I would suspect that the 
electron density clarity would not be terrific.

5 molecules in the ASU seems a bit odd to me. (4 or 6 seems more 
probable.) If it is 5 molecules, one of the molecules would have to be 
on a symmetry axis, completing a "symmetry" dimer, with two other dimers 
in the ASU? I've seen 6 molecules in the ASU in C2, with one tetramer 
and one half-tetramer on the symmetry axis. When solving that one, the 
Matthews Probability calculator insisted that there were 4 molecules in 
the ASU. Turned out to be 6. I initially placed 4 molecules  in the ASU, 
and saw a clear crystal-packing gap for another dimer that would fit 
neatly next to the symmetry axis.

Roger S. Rowlett
Gordon & Dorothy Kline Professor
Department of Chemistry
Colgate University
13 Oak Drive
Hamilton, NY 13346

tel: (315)-228-7245
ofc: (315)-228-7395
fax: (315)-228-7935
email: rrowlett at colgate.edu

On 12/12/2016 9:52 AM, oleg at quantumbioinc.com wrote:
> Compute Matthews coefficient.
> e.g. using
> http://www.ruppweb.org/mattprob/default.html
> Oleg
> -----Original Message-----
> From: "Sharan Karade" <sharankarade at gmail.com>
> Sent: Sunday, December 11, 2016 6:05am
> To: phenixbb at phenix-online.org
> Subject: [phenixbb] Asymmetric unit content
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb
> Unsubscribe: phenixbb-leave at phenix-online.orgDear all,
>             I have solved the structure in C121 space group, having five
> molecules in asymmetric unit, my supervisor have doubt about the fifth
> molecule becoz of poor density fit, if i delete the fifth molecule, there
> is a space for the molecule and crystal is not continuous, i calculated
> difference map from Phenix by deleting fifth molecule and got the Map
> showing presence of molecule. My supervisor want me to get SetMet data, but
> crystals are diffracting poor (7A). Is there any way to convince my guide.
> advance thanks for valuable suggestions.

More information about the phenixbb mailing list