[gmx-users] Umbrella sampling - energy calculations

[Александр Журавлев]

Dear Gromacs users! I wonder does it exist somewhere a full description (in manual, tutorial...) how to calculate free energy using Umbrella sampling, for example, for ligand-receptor dissociation? I read the manual but I am not sure that I understood it completely.
Is it necessary to specify coordinates of the ligand position outside the receptor binding pocket (to which the dissociation should lead)? 
How is it possible to specify the correct direction of ligand movement?
Should I make one md stimulation of the system for one ligand position or several for different lambda (from 0 to 1) that describes the changing of its position? Or maybe this refers only to TI method?
Thanks,

Alexander