defining states in Gibbs

From: Sophia Kondratova <>
Date: Wed 22 Aug 2001 16:39:35 -0300

I am just starting to work with AMBER and I am a bit confused with defining
the two states that will be compared by the Gibbs module. Is this correct? In
simulation of X -> Y

State 1 = the initial state of the system (lambda = 1) X
State 0 = final state of the system, the perturbed state ( lambda = 0) Y
state 1 is X and state 0 is Y, so perturbation occurs from lambda=1
->lambda=0. Usually in the literature the perturbation occurs from 0->1, and
the free energy is found by G = G1- G0. In the case where the perturbation
occurs from 1->0, wouldn't this be correct: G = G0 - G1.

Also, if the standard "forward" sampling is 0->1, if what I understand is
correct where 1 is the initial state and 0 is the final state, wouldn't than
mean that you would be calculating (final state)0 -> (initial state)1 ( that
does not look like forward sampling, but more like reverse sampling)

Any clarification would be greatly appreciated.

Sophia Kondratova
Received on Wed Aug 22 2001 - 12:39:35 PDT
Custom Search