On Thu, Sep 1, 2011 at 8:07 AM, jens j birktoft
What would be proper approach to generate restraints that will keep base pairs coplanar (or nearly coplanar) in RNA/DNA helical structures? Hydrogen bonds by themselves are not sufficient, at least not for low resolution refinement, <4Å
The only way to do this right now is to define custom planarity restraints, like this: refinement.geometry_restraints.edits { planarity { action= *add delete change atom_selection= (chain A and resseq 1 and (name X or name Y or name Z or ...)) or (chain B and resseq 10 and (name X or ....)) sigma= 0.1 } } Increasing sigma will make it looser, but I'm not sure what the scale is. It will probably need to be relatively high to allow for propeller twist (or other deformations - I've long since forgotten what these are); the individual base planarity restraints will continue to keep those very flat and rigid. We've discussed adding this restraint internally, but we haven't decided the best way to implement it - I suppose we could at least streamline the definition of custom planes. However, we are short on good test cases, so if anyone has an example of a *small* structure that needs the additional restraints, please contact me off-list. -Nat