You will have to use, for example, the weighted histograms analysis method (the so-called WHAM). If you have run K replicas distributed among L thermostats with

inverse temperature \beta_l (l= 1, 2, . . . , L) and with potential energy U_l(q _k) (l = 1, 2, . . . , L; k = 1, 2, . . . , K). You can then histogram the potential energy into bins, U_m (m = 1, 2, . . . , M), combining all of the replicas. Thus, for each

replica k, a histogram is constructed and the number of

independent snapshots in the potential energy bin

m, N_{km} can be accumulated. The probability of finding the system in the bin m

with energy U_m at the inverse temperature β_l can be written

 P_{lm}= Z^{-1}_l C_{lm} \Omega_m exp( −\beta_0 U_m),

where \Omega_m = \Omega(U_m ) is the density of states,

\Omega_m exp( −\beta_0 U_m)

represents the unbiased probability of the bin m at the

reference inverse temperature \beta_0 , and C_{lm} is the bias factor

that accounts for the effect of the temperature:

C_{lm} = exp[-(\beta_l - \beta_0) U_m]

Calculations of Z^{-1}_l and \Omega_m can be performed using the scheme provided in references:

1) E. Gallicchio, M. Andrec, A. K. Felts, and R. M. Levy, J. Phys. Chem. B

109, 6722–6731 (2005).

2) J. D. Chodera, W. C. Swope, J. W. Pitera, C. Seok, and K. A. Dill, J. Chem.

Theory Comput. 3, 26–41 (2007).

Then, the average of any physical property can be evaluated as

= \sum_k \sum_n W_{kn}  A_{kn}  / \sum_k \sum_n W_{kn}

with Boltzmann's weights:

W_{kn} = \sum_m [\Omega_m/N_{km}] exp(-\beta_0 U_m)

index n runs over all snapshots of replica k and k runs over all replicas.

Thank you very much for your reply. I am trying to do the analysis accordingly.

http://www.drugdesign.gr/uploads/7/6/0/2/7602318/lecture_mdanalysis.pdf

Dear Kalyanashis, How did you use wham for the remd?  please advice me.

Sorry.. I did completed it. Please ask to other expert.