Mainly we discussed the problems of minimization. In the input script
for SANDER the dielectric constant was wrong. Minimization with GB may
be useful?
Docking of ligands into the active site without water molecules is
leading to wrong conformations of the ligand.