On Fri, Apr 18, 2014 at 6:53 AM, 李翔 <lixiang1642@gmail.com> wrote:

I am working my low resolution data (6A) by MR and met a problem about DNA basepairing during refinement. I think so far everything looks OK except the bases flipped during the refinement. I added basepairing restraint for all bases, however, seems not help too much. Does anyone have experience on that?

First of all, make sure you have real-space refinement turned off; this may be moving the DNA too aggressively. Second, it sounds like the base-pairing restraints may not be working correctly. If you send me the model (off-list!) I can take a look and see if there is a bug. The other option is to restrain the coordinates to the starting structure - I think the reference model restraints would be best but I'm not entirely sure if they work for DNA. (If not, you can use harmonic restraints, but these are less flexible.)

-Nat