Re: [phenixbb] phenix.ready_set

Hi, just to second this... In refinement hydrogens are treated as "riding" (unless you are refining against neutron or ultra-high resolution X-ray data). Since Reduce (or ReadySet! that uses it) builds hydrogen atoms in a smart way (optimizes hydrogen bonds, for example) it's a good idea to rebuild hydrogens from time to time because it may build better hydrogens starting form a better model. I was even thinking to rebuild hydrogens during refinement every macro-cycle by calling Reduce internally but I haven't done it yet because of numerous technical troubles, including inconsistencies in geometry libraries used by Reduce and Monomer Library, different naming conventions, etc etc etc. Pavel On 2/10/12 8:36 AM, Jianghai Zhu wrote:

Exactly my point. The structure actually changes a lot during building, especially for low-resolution structures. Also another thing about keeping the hydrogen is that COOT runs very slow in the presence of hydrogens. And regularization with hydrogen in COOT often gives different result than without hydrogen.

-- Jianghai

On Feb 9, 2012, at 7:21 PM, Nathaniel Echols wrote:

...

On Thu, Feb 9, 2012 at 4:16 PM, Nigel Moriarty wrote:

...

Its true that ReadySet! does not handle the glycans as well as it should. Its on my list of things to fix, but you can delete the hydrogens by hand for a short time. This actually brings up an interesting point. I believe its better to add the hydrogens once and use that model for all subsequent refinements. In this scenario, you use ReadySet! once and the addition of hydrogens is a "checkpoint" in the refinement process. This only makes sense if you assume that no more atoms which might contain hydrogens will be added during refinement - which is very rarely the case until the very end. It would be helpful if ReadySet didn't add hydrogens to ligands that already have one or more built, since this will both avoid the replacement of spurious hydrogens on glycans and also make subsequent runs much faster.

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

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