# Number of replicas in the REST2

GROMACS version:
GROMACS modification: Yes/No

Dear all
I am running REST simulation for a system that has ( Protein, DNA, and RNA) sing 40 replicas over a temperature range (300-500K)
I did a test of 2 ns.

I have some Qs
1- are the number of replicas too much? I can run 3ns/day for each replica, so the total is 120 ns/day for the 40 replicas.

2- I analyzed the PE overlap between every two replicas; there is a good overlap between each two which is a good sign. But I am concerned about the exchange probabilities.
To obtain sufficient potential energy overlaps between the two replicas, the acceptance ratio between them should be 0.2~0.3 is good.
Are these values good enough?

Repl average probabilities:
Repl 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
Repl .09 .06 .09 .10 .10 .10 .10 .11 .12 .09 .13 .13 .14 .14 .13 .14 .15 .13 .17 .16 .15 .16 .17 .18 .19 .18 .19 .19 .20 .20 .20 .20 .21 .20 .21 .22 .22 .22 .23
Repl number of exchanges:
Repl 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
Repl 421 297 452 487 516 488 520 525 637 449 641 654 697 657 634 700 761 659 830 804 758 798 858 888 926 879 967 984 1000 977 976 1039 1071 1015 1065 1116 1102 1103 1140
Repl average number of exchanges:
Repl 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
Repl .08 .06 .09 .10 .10 .10 .10 .11 .13 .09 .13 .13 .14 .13 .13 .14 .15 .13 .17 .16 .15 .16 .17 .18 .19 .18 .19 .20 .20 .20 .20 .21 .21 .20 .21 .22 .22 .22 .23

3- -replex = Instruct the system to attend an exchange at every 100 steps. Do I need to increase this number to 500 or 750?

Any response or feedback is highly appreciated.
Thank you so much