Fig. 3. A detailed investigation of the p38α protein–ligand complexes. The apo (1wfc,53 orange) and holo (3fly, blue) structures have several structural differences close to the ligand binding site: a substantial loop motion and a different T106 rotamer state. In the simulations, the rotamer T106 retains its initial state: shown in lines, with a sphere marking threonine's oxygen. The calculated ΔG values depend strongly on the starting structure (holo or apo) that is used to initialize protein simulations in its apo state: scatterplots at the bottom. The green structure in the sub-panel and corresponding ΔG scatterplot depict a case, where apo simulations were initialized with a holo structure (ligand removed), but with the T106 rotamer set into its apo state. Dark and light shaded areas represent regions deviating from experiment by at most 1 and 2 kcal mol−1.
