Average running time, running time fold improvement, and memory usage of optimized ZDOCK versions.

All values are averages from running ZDOCK on 176 unbound docking test cases, each run using a single 2.8 GHz 64-bit Opteron processor with 8 GB available RAM. Bold rows correspond to the fully optimized ZDOCK versions. 1Optimization scheme. Conv3D = new 3D FFT library, Cent = optimal receptor centering, Rot = optimal receptor rotation, Switch = switch ligand and receptor. See Implementation section for details. 2Average fold improvement in running time versus the previous major ZDOCK version (3.0 or 2.3).