Re: [phenixbb] DEN refinement using CC instead of Rfree

Nat is correct, torsion dynamics at present can only be run on the entire structure. He's also correct that using a selection won't speed up the process, I meant to mention that in my original message, so thanks for clarifying. Jeff On Fri, Oct 25, 2013 at 10:21 AM, Nathaniel Echols wrote:

On Fri, Oct 25, 2013 at 7:04 AM, Jeff Headd wrote:

...

In theory you should be able to run DEN on a selection, but in my own simple test this did not work. I'll see if there is a bug that can be fixed.

I may be mistaken, but I think the torsion dynamics can only be run on the entire structure - there is an atom selection parameter for it in phenix.refine, but this doesn't actually do anything.

It wouldn't make the program faster anyway, because it's refining against reciprocal space data, and the FFT is the main bottleneck. It would be interesting to try doing this in real space, though.

-Nat

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