[gmx-users] Re: Hamiltonian replica exchange umbrella sampling with gmx 4.6

[Dejun Lin]

Hi Micheal,

Just to make sure I understand you correctly -- I can have 1 type of
restraint, e.g., a harmonic potential between 2 groups defined by (k1, d1)
where k1 is force constant and d1 is the equilibrium position. But I can't
have 2 or more types of restraints like (k1,d1,k2,d2,...,kn,dn). Is that
correct? 

Also, about to paper you referred to. Is the algorithm described in section
IIIB-2 the one that's currently implemented in gromacs? If so, how do you
"precompute" the matrix U for all pairs without knowing x? I guess I just
don't have an idea about how the whole procedure is conducted. 

Thanks,
Dejun



--
View this message in context: http://gromacs.5086.n6.nabble.com/Hamiltonian-replica-exchange-umbrella-sampling-with-gmx-4-6-tp5006187p5006853.html
Sent from the GROMACS Users Forum mailing list archive at Nabble.com.