Chapman Group
Contact Information
Biomolecular Structure Laboratory
CV (full)
(NIH biosketch)
Graduate studies
People (trainees)
Software Distribution & Documentation
Page Links
Graduate programs
Other Links
Dept. Biochemistry & Molecular Biology
Portland Biophysics Society



For the fitting of atomic models into density maps derived from x-ray crystallography or electron microscopy (EM). Can be used in two modes:

  • Stand-alone program for unrestrained real-space rigid-group and/or restrained B-factor refinement and for optimization of EM imaging parameters vs. an atomic model, all by least-squares gradient-descent.
  • Extension module for other optimizers (eg. CNS) enabling stereochemically-restrained real-space refinement by gradient descent or simulated annealing using Cartesian or torsion angle coordinate systems.

This is a completely new implementation described in: Chapman, M. S., Trzynka, A., and Chapman, B. K. (2013) Atomic modeling of cryo-electron microscopy reconstructions - Joint refinement of model and imaging parameters, J Struct Biol 182, 10-21.
This is an extension of an approach first described in Chapman, M. S. (1995) Acta Crystallographica A51, 69-80 (1995).

RSRef is documented and distributed as part of the PaStO package, see below.


For the superimposition of protein structures, moving and target. The pairing of atoms is through user-specified combinations of order and type. The moving molecule can be treated as one or more rigid groups, and/or flexibly with variable backbone dihedral angles throughout or in selected regions, optionally adding a parsimony restraint for minimalist changes. This allows the characterization of hinge regions from pairs of crystal structures, the method having been cross-checked against NMR measures of dynamics.

The approach is described in Chapman, B.K., Davulcu, O, Skalicky, J.J., Bruschweiler, R.P. & Chapman, M.S. (2015) Parsimony in Protein Conformational Change, Structure, in press;

Superpose is documented and distributed as part of the PaStO package, see below.

PaStO - Parsimonious Structure Optimization:

This is a package of programs that share a need for optimization of protein structures using models with optionally reduced parameterization to minimize overfitting, particularly when working with data of limited resolution.


  • Installation / new user README.txt.
  • On-line documentation.
  • Documentation is also distributed with the package (as web pages or a single pdf) for local access.

Download (version: 0.5.6; August 2016)

Licensing (free for academic use) can be initiated at Upon receipt, instructions will be provided for accessing the distribution. RSRef is coded in Python and is compatible with diverse platforms.

Right click and save target: Encrypted tar distribution; (Instructions for decryption, when you have the password.)

PaStO will soon also be available through SBGrid.


Pre-processor for the subtraction of ice rings from diffraction images (Chapman, M. S. & Somasundaram (2010), T. De-icing: recovery of diffraction intensities in the presence of ice rings. Acta Crystallogr D Biol Crystallogr 66, 741-744,  A self-documenting Python script can be saved (, together with example control input (run), input data (AKAp_1_001.img) and output (di_AKAp_1_001.img)
. is currently configured for ADSC Quantum 4 & like detectors, but is extendable to others.  For full documentation use "pydoc deice" and " -h".  Imports modules from NumPy.


The following programs are no longer actively supported, but continue to be available as-is from Chapman's former institution, Florida State University.


A program to display the surface of a macromolecule, and its properties. (UNIX).


Module for refinement of macromolecular structures using orientational data from solid-state NMR. (UNIX).


Module for including an angle-dependent hydrogen bond restraint in crystallographic refinement of macromolecules. (UNIX and Linux).


Module for including an electrostatic Poisson-Boltzmann restraint in crystallographic refinement of macromolecules. (Linux).


Module for calculating less bias R-free by excluding neighboring reflections of test set reflections based on G-function. (Linux).

Last updated: 06/09/15