Figure 3.

Results of 1000 iteration minimization for the CABS decoys. For each protein, the energy was plotted as a function of Cα RMSD for all decoys (left panels) and without decoys with abnormal high energy values resulted from structural inaccuracies (right panels). On the left panels, energies of the native structures are denoted by asterisks. The native structures were subjected to the same rebuilding procedure from the Cα-traces as that applied to the decoys. Proteins are ordered in respect to their chain lengths (Table 1) – from the smallest on top (2GR8A) to the largest (2CJPA) on the bottom.