Figure 1.

ARCIMBOLDO operated workflow starting from fragment subsets as initial molecular-replacement models assigned to Phaser, which performs a rotation and translation search including a subsequent refinement. Depending on the ARCIMBOLDO setup, all molecular-replacement results or results better than a specified average will be passed automatically to SHELXE. After subsequent and iterative density modification and auto-tracing, successful SHELXE expansion results could be identified by sorting the SHELXE CC (correlation coefficient) values. In our case of protein–DNA targets, CC values above 20% tagged a successful solution for a specific PDB start fragment.