Thanks Guys,
 
would it be possible to implement the H-bonding criteria? From my experience with CNS it worked quite nicely. Right now I end up with ~350 waters in a 2.3 A structure of a 30kDa protein, which is a bit much.
-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Monday, March 19, 2007 1:02 PM
To: PHENIX user mailing list
Subject: Re: [phenixbb] More stringent water picking

Hi Carsten,

thanks for your question!

The main water picking criteria in phenix.refine are:
1) Fo-Fc map;
2) 2Fo-Fc map;
3) macromolecule - water distance;
4) water-water distance;
5) B-factors and occupancies  max/min values;
plus some more of expert level ...

There is no specific H-bonding criteria implemented.  You can play with sigma cutoff for both maps and with the distances above. In most of the cases it the reasonable behavior.

I would be interested in Paul's (Adams) comments about his experience of usefulness of H-bonding criteria in CNS.

Thanks!
Pavel.



Schubert, Carsten [PRDUS] wrote:

Hi,

is it possible to increase the stringency criteria for picked waters, besides increasing the sigma level? E.G. something on the order of the H-bonding criterion found in CNS? I find that the automated water-picking routine gives me quite a few waters, which are not necessarily justified by visual inspection of the maps.

Thanks

        Carsten 



_______________________________________________
phenixbb mailing list
[email protected]
http://www.phenix-online.org/mailman/listinfo/phenixbb