Here one can find the topology files used in the paper

ligand*/complex/ contains the complex topology

ligand*/complex_boresch/ contains the complex topology with the boresch restraint (only ligand 1 to 6) (Same as orginal paper http://dx.doi.org/10.1039/C5SC02678D)

ligand*/complex_LT/ contains the complex topology with one ligand atom restained (it has to be used during the alchemical transformations when using the Loose-Tight restraint, but not during the HREM)

ligand*/complex_LT_twoatoms/ contains the complex topology with two ligand atoms restained (it has to be used during the alchemical transformations when using the Loose-Tight restraint, but not during the HREM) (only ligand 1 to 6)

ligand*/ligand_in_vacuum/ contains the topology of the ligand in vacuum

ligand*/ligand_in_water/ contains the topology of water + ligand (in this order)

In ligand 6, 7, and 8 there are also directories with the _new suffix, those contain the topology where the dihedrals were reparametrized with the ANI-2.X forcefield throught the playmolecule web interface, while the rest of the topology was left untouched

NOTE:
In the original paper the ligand and the protein were written as one [ moleculetype ]  while here the ligand has the LIG [ moleculetype ]
I did this by using the python library called parmed (https://github.com/ParmEd/ParmEd)

$ python
$ import parmed
$ pp = parmed.load_file('complex_with_dummy.top', xyz='complex_with_dummy.gro')
$ (pp['!:LIG'] + pp[':LIG']).save('new.top')

and then changed manually the order of the molecules in the [ molecules ] section of the topology

