[gmx-users] REMD ... again a problem with analysis (on the fly)

[mollica.luca at hsr.it]

Hi,

unfortunately, waiting for the end of REMD simulation, I can 
only  ‘grep’ my log files for each replica run in order to check 
what’s going on … say, md0.log, md1.log …. up to md25.log .

What I can see in each of my log file is (every 2 ps or 1000 steps, 
having set this value as exchange rate) something like this (this is 
extracted form replica 3):


Replica exchange at step 48000 time 96
Repl 3 <-> 4  dE =  4.036e+01  dpV =  1.591e-03  d =  4.037e+01
Repl ex  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
Repl 
pr        .00       .00       .00       .00       .00       .00       .
00       .00       .00       .00       .00       .00     


Replica exchange at step 49000 time 98
Repl 2 <-> 3  dE =  5.380e+01  dpV =  9.683e-04  d =  5.380e+01
Repl ex  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
Repl 
pr   .00       .00       .00       .00       .00       .00       .00   
    .00       .00       .00       .00       .00       .00


Actually, I do not get the point of what is going to happen: does the 
statement 

Repl 3 <-> 4  dE =  4.036e+01  dpV =  1.591e-03  d =  4.037e+01

mean that replica 3 and 4 are going to exchange ? This would mean in 
my case that a replica exchange event is happening every 2ps, but this 
seems to happen for every replica and it seems so strange also due to 
the fact that the probability for each exchange is set to .00 for 
every replica couple for each attempt ….
Do I have to wait for the end of simulation and get the final 
statistics of the job ? Or am I missing completely the point ? If yes, 
are going to be provided any detailed explanations of the REMD 
features of GMX 3.3 in the near future ?

Thanx a lot and kind regards


LM