A number of users find that autoBUSTER gives more molprobity C-beta deviations than their software of choice. See BUSTER FAQ for a discussion. This tutorial shows how to use restraints to reduce the effect.

We'll start with a pretty much finished refinement from a different example AutoBusterExample1osgExtraPeptideTLS: 1osg.

You have the 1osgBasicNCS_004_TLSfrom_003_refine.pdb and 1osgBasicNCS_004_TLSfrom_003_refine.mtz files available.

Upload 1osgBasicNCS_004_TLSfrom_003_refine.pdb to the molprobity site at http://molprobity.biochem.duke.edu

Click 'add hydrogens' to get the interim report mp-flipreview.png which indicates that ASN 235 is probably the wrong way round in all six chains. Accept the flips, save the reduced PDB file, and proceed to 'Analyze all-atom contacts'. Keep the default set of actions and reports, and let 'er rip; the final report is mp-report.png

Directly restraining C-alpha chirality

To re-refine with C-alpha chirality restrained (which we do using an improper torsion term with a rather tight sigma of 1.7 degrees), you need to add an additional restraint file, which can be collected from


This is quite a large structure and the run (with the command below) takes 2600 seconds on a fast quad-core PC

refine -p 1osgBasicNCS_004_TLSfrom_003_refine.pdb \
       -m 1osg.mtz -autoncs -M TLSbasic \
       -Gelly protgeo_option_chiralrestraint_from_equilib.dat \
       -d add-chirest | tee add-chirest.lis

so the results are available as add-chirest.pdb and add-chirest.mtz

Upload add-chirest.pdb to the molprobity site and continue as before. The interim report


indicates that I didn't flip ASN 235 before the re-refinement; the final report


shows that we've got rid of twenty C-beta deviations.