Macromolecular structure refinement with phenix.refine
Pavel Afonine

The talk will overview the recent advances and available features of phenix.refine – a general purpose state of the art refinement module of PHENIX. New or improved crystallographic algorithms, implemented using highly efficient programming tools, can provide a very high level of automation and robustness during structure refinement.

Key features of phenix.refine include:

  • Automatic handling of most data formats
  • Individual coordinates refinement using minimization or simulated annealing
  • Highly optimized and automated rigid body refinement
  • Atomic displacement parameter (ADP) refinement: individual isotropic or anisotropic, grouped, TLS or any mixture
  • Occupancy refinement (grouped, individual, constrained for alternative conformations)
  • Automatic NCS detection and use in refinement as restraints
  • Automatic twinning detection and use in refinement
  • Various refinement targets (maximum-likelihood, maximum-likelihood with experimental phase information, and amplitude least-squares)
  • Robust mask-based bulk-solvent correction and anisotropic scaling
  • sigmaA-weighted map output
  • Refinement using X-ray data, neutron data or joint refinement with both
  • Refinement at ultra-high resolution using an Inter Atomic Scattering model

Multiple application examples will be presented as well as some general introduction to phenix.refine usage.