Hi Kendall, No, ligandfit finds the core of your ligand, then traces the remainder, rotating torsion angles for the ligand using a simple algorithm. It does this several times, and the pieces are stitched together. No refinement is done. This can yield poor geometry as you found. The best solution is to then refine the model+ligand together. Usually that will take care of the poor ligand geometry. All the best, Tom T

...

I just tried ligandfit for the first time and it did not maintain proper geometry of the ligand. There were problems with both bond lengths and planarity. Is there a way to specify a .cif file? The ligand pdb was generated with PROGRD2, and I have had good success refining a structure that contains a very close analog with phenix.refine.

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