Fwd: Bug in T-REMD with colvars

From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Tue Dec 03 2013 - 13:31:24 CST

Should replica.namd be hardcoded into namd2.10 (which I used for the above
report), I had no chance, and the failure I was faced with might really be
a bug of namd2.10. I forgot to add that in previous mail.
francesco pietra

---------- Forwarded message ----------
From: Francesco Pietra <chiendarret_at_gmail.com>
Date: Tue, Dec 3, 2013 at 9:55 AM
Subject: Fwd: Bug in T-REMD with colvars
To: NAMD <namd-l_at_ks.uiuc.edu>

ADDENDUM: to let verifying whether the line I added to replica.namd is
correct, I add here the configuration of the launching file:

*****************************
#!/bin/bash
# @ job_type = bluegene
# @ job_name = P2i
# @ wall_clock_limit = 00:20:00
# @ output = $(job_name).$(jobid).out
# @ error = $(job_name).err
# @ shell = /bin/bash
# @ bg_size = 64
# @ account_no = xxxxxxxxx
# @ environment = COPY_ALL
# @ notification = always
# @ queue

# module load namd/2.10 REMD
module load profile/advanced
module load namd/2.10

exe=$(which namd2)
# Setup replica exchange
mkdir output
cd output
mkdir 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
cd ../

# Running on 64 nodes with 32 tasks per node (2048 tasks in total)
# runjob -n 2048 -p 32 : $exe +replicas 8 job0.conf +stdout
output/%d/job0.%d.log

runjob -n 2048 -p 32 : $exe +replicas 32 job1.conf +stdout
output/%d/job1.%d.log

*********************************

In replica /0, as an example for all 32 replicas;

 fold_P2i.job0.0.colvars.state
fold_P2i.job0.0.colvars.traj
fold_P2i.job0.0.dcd
fold_P2i.job0.0.history
fold_P2i.job0.restart500.0.coor
fold_P2i.job0.restart500.0.tcl
fold_P2i_job0.restart500.0.vel
fold_P2i.job0.restart500.0.xsc
fold_P2i.job1.0.colvars.traj
fold_P2i.job1.0.history
job0.0.log
job1.0.log

francesco pietra

---------- Forwarded message ----------
From: Francesco Pietra <chiendarret_at_gmail.com>
Date: Tue, Dec 3, 2013 at 8:55 AM
Subject: Bug in T-REMD with colvars
To: NAMD <namd-l_at_ks.uiuc.edu>

Hello:
Bug in T-REMD with colvars could hopefully simply mean bug in information
about T-REMD with colvars.

Some time ago I asked - on general grounds - advice on how to manage to get
colvars state be recognized by restarted replicas. I got no answer. Here, I
am more specific on this issue, working with namd2.10, although same issue
with namd2.9.

Setting in namd config file (alanin_based.namd in the alanin t-remd example
provided by namd)

# constraints
colvars on ; # provide colvarsConfig (and colvarsINput if
restart)
colvarsConfig ./P2i_rmsd-colvars.in

gives correct
fold_P2i.job0.0.colvars.state
fold_P2i.job0.0.colvars.traj

in replica/0, like in all other 31 replicas of my T-REMD. However, what is
not provided anywhere is what the system asks (see below, please) when I
add a line for colvarsInput into file replica.namd as follows:

if {$i_run} { #restart
  bincoordinates $restart_root.$replica_id.coor
  binvelocities $restart_root.$replica_id.vel
  extendedSystem $restart_root.$replica_id.xsc
  colvarsInput $restart_root.$replica_id.colvars.state
} else {
  temperature $NEWTEMP
}

Thus, the error arises in the .err file:

REPLICA 23 FATAL ERROR: Error in the collective variables module: exiting.

Looking at the replica 23 log file, one learns that:

colvars: Collective variables biases initialized, 1 in
total.fold_P2i.job0.restart500.23.colvars.state
colvars:
----------------------------------------------------------------------
colvars: Error: in opening restart file
"output/23/fold_P2i.job0.restart500.23.colvars.state".
*******************************

In this scenario, I wish that the line "colvarsInput
$restart_root.$replica_id.colvars.state" I added to replica.namd should be
reformulated, or placed elsewhere (in the namd config file?). That
statement was introduced on imitation of umbrella sampling.

I hope now that developers will be so kind to correct my faults or my
overlooking any T-REMD output, if any, to the benefit of all namd users
too. Or provide anyway a solution to carry out correctly T-REMD with
colvars.

Thanks a lot for advice

francesco pietra

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:24:03 CST