Can modeller do ligand-steered modelling? And what is the difference
between the ligand-steered method and the mothods from the on-line
tutorial of Modeling ligands in the binding site? Are they the same
method? Thanks very much!
Ligands can be handled in Modeller in two ways: either as flexible
residues (if topology and parameters are present in the CHARMM
forcefield files, modlib/top_heav.lib and modlib/par.lib) or as 'block'
residues, where they are simply copied from template to target as rigid
bodies. The examples in the tutorial use the latter method. Neither of
these qualify as "ligand-steered" if by that you mean refinement of
shape of the binding pocket - binding residues will be treated much like
any other protein residue, and so will tend to look like the binding
pocket in the template.