[gmx-users] Re: forward and reverse free energy not consistant (Li Qiang) (Diana Lousa)

[ggroenh]

To rule out a sampling problem you can perform the simulations with  
everything else frozen or position restraints. Both backwards and  
forwards, and you can do it very fast, as there is nothing to  
equilibrate. If there is still a discrepancy between the two paths,  
you should suspect a problem with your topology, with the bonded  
interactions for instance.

Gerrit