I'm now doing paired refinement with and without real space refinement and will send you the files when they are done. Thanks, Owen On Jul 15, 2013, at 3:03 PM, Pavel Afonine wrote:
Hi Owen,
local real-space refinement will not fit a side-chain into a rotamer outlier: it uses both, map and rotamer library, and also it uses a quite complex logic that aims to find best density fit and at the same time keep good geometry (no outliers of any sort). That's why I wanted to look at the files because to me what you described is an indicator of a problem somewhere else or a bug..
Pavel
On 7/15/13 11:56 AM, Owen Pornillos wrote:
Hi Pavel -
Thanks for the reply.
I have a 2.6 A data set that has a dimer in the asymmetric unit, and so I'm using NCS. Some of the big sidechain densities (LYS, ARG, etc.) are not well-defined so I'd like to just restrain them to the other chain. The auto rotamer fitting of course tries to fit them, and the result is a list of half a dozen or so rotamer violations. I'm going through the list to see if the density truly defines different conformations. I think this is just a general run of the mill problem with fitting/refinement with NCS.
Thanks,
Owen
On Jul 15, 2013, at 1:55 PM, Pavel Afonine wrote:
Hi Owen,
you can disable it by turning of real-space refinement off.
I'm interested to know what exactly is not going well: could you please explain this to me off list and send files so I can reproduce the problem?
Pavel
On 7/15/13 10:46 AM, Owen Pornillos wrote:
I'm using one of the nightlies - dev 1427. It's not doing very well in combination with NCS. Or what is the most recent nightly that does not have this feature?
Many thanks,
Owen
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb