Chemistry Reference
In-Depth Information
In the Replica exchange TI method (RTI), hamiltonian replica exchange moves
are incorporated between adjacent λ simulations. This becomes periodic yielding
exchanges between each configuration and its neighbor. Moves are rejected or
accepted according to metropolis tests. Sampling is improved and simulations are
able to move freely to configurations that are more favorable [587].
In the procedures of Adaptive umbrella sampling and adaptive umbrella weighted
histogram analysis methods (ADuMWHAM) the parameter λ is treated as a
dynamic coordinate which is moved throughout the simulation. The relative free
energy is obtained from the potential of mean force along λ [317, 602].
In the parallel tempering (PT) method, multiple replica of the system is used to
enhance sampling. Each replica runs at different temperatures. A Monte Carlo test
is made with neighboring temperatures which can result in changes of the system'
coordinates at these temperatures. Finally, a correct ensemble distribution is
obtained from the set of trajectories at each temperature. The approach of PPTI
(parallel tempering thermodynamic integration) uses different temperatures with
replicas of the system. In this method, each replica runs at a different temperature
[602].
Based on linear response theory, LIE is a linear interaction method. In this
procedure, the compound's binding free energy is determined from the average of
the potential interaction energy ensemble. The approach can use semi empirical,
molecular dynamics and electrostatic models to estimate free energies between
protein and ligands [604]. It can also be used as a scoring function.
Another approach to simulate free energies is the one-step (OS) perturbation
method. In this procedure, rather than connecting A and B states by intermediate
simulations, a single reference state does the work of sampling the degrees of
freedom for both A and B states. Considerable gain is obtained if one state is
chosen and used to calculate free energies for a series of end states.
The approach (LIE/OS) results from a combination of LIE and OS. One
advantage of the combination is that the apolar contribution to the free energy is
improved by the OS participation. Simulations of physically relevant states can
