Plumed2.1-hrex with Gromacs5.1

467 views
Skip to first unread message

nicolas...@gmail.com

unread,
Nov 4, 2015, 4:49:15 AM11/4/15
to PLUMED users
Dear all,

I am trying to use Plumed2.1-hrex with Gromacs5.1. I have successfully installed plumed, patched Gromacs and compiled it. Then I prepared my different topologies with plumed-partial_tempering and the tpr with grompp. I want to do HREX, so I used the following command:
mpirun -np 4 /home2/ncheron/Programs/Gromacs_5.1/bin/gmx_mpi_d mdrun -plumed plumed.dat -multi 4 -replex 1000 
 
In the beginning of the md0.log file, I have:
*****************************************
PLUMED: PLUMED is starting
PLUMED: Version: 2.1.3 (git: Unknown) compiled on Nov  3 2015 at 18:06:12
PLUMED: Please cite this paper when using PLUMED [1]
PLUMED: For further information see the PLUMED web page at http://www.plumed-code.org
PLUMED: Molecular dynamics engine: gromacs
PLUMED: Precision of reals: 8
PLUMED: Running over 1 node
PLUMED: Number of atoms: 16826
PLUMED: GROMACS-like replica exchange is on
PLUMED: File suffix: .0
PLUMED: FILE: plumed.dat
PLUMED: END FILE: plumed.dat
PLUMED: Timestep: 0.001000
PLUMED: KbT: 2.494339
PLUMED: Relevant bibliography:
PLUMED:   [1] Tribello, Bonomi, Branduardi, Camilloni, and Bussi, Comput. Phys. Commun. 185, 604 (2014)
PLUMED: Please read and cite where appropriate!
PLUMED: Finished setup
*****************************************
which seems to mean that plumed is working and is loaded.

However, HREX doesn't seem to work: everytime an exchange is attempted, I have the following:
*****************************************
Replica exchange at step 5000 time 5.00000
Repl 0 <-> 1  dE_term = -0.000e+00 (kT)
dplumed =  0.000e+00  dE_Term =  0.000e+00 (kT)
Repl ex  0 x  1    2 x  3
Repl pr   1.0       1.0
*****************************************

I know that for Gromacs4.6.7, there was an additional flag "-hrex" for mdrun but it doesn't work anymore with Gromacs5.1. Am I missing something? Is it possible to perform HREX with Gromacs5.1?

Thank you for your help.

Nicolas

Giovanni Bussi

unread,
Nov 4, 2015, 8:17:53 AM11/4/15
to plumed...@googlegroups.com
Hi,

I am sorry but I didn't have time to port it to gromacs 5.

Perhaps I will do it in the future.

Giovanni

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To post to this group, send email to plumed...@googlegroups.com.
Visit this group at http://groups.google.com/group/plumed-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/8a407947-5064-4ca2-a4e5-1a8c33983e9a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Chris Neale

unread,
Mar 29, 2016, 6:26:16 PM3/29/16
to PLUMED users
I am considering doing this porting. Any idea how much work it would be? (days vs. months?)

Thank you,
Chris.

Giovanni Bussi

unread,
Mar 30, 2016, 2:33:26 AM3/30/16
to plumed...@googlegroups.com
Hi Chris,

I have a tentative implementation already. The reason why I do not share it is that none in my group is presently using gromacs 5, so I am afraid I would not be able to properly test it. If you are willing to make some test I would be happy to put it on github.

Giovanni

Davide Provasi

unread,
Mar 30, 2016, 7:42:12 AM3/30/16
to plumed...@googlegroups.com, Sebastian Schneider
Hi Giovanni and Chris,
we'd be also glad to contribute testing a hrex port to gmx5.

Davide



For more options, visit https://groups.google.com/d/optout.


Davide Provasi
Dept. of Structural and Chemical Biology
Mount Sinai School of Medicine
Icahn Medical Institute Building
1425 Madison Avenue, Box 1677
New York, NY 10029-6574
Tel.:212-659-8618
Fax: 212-849-2456

Giovanni Bussi

unread,
Mar 30, 2016, 9:04:26 AM3/30/16
to plumed...@googlegroups.com, Sebastian Schneider
Hi all,

I uploaded a new branch here:

This is based on plumed 2.2, top of branch. As usual, patches are expected to be compatible across plumed versions.

This branch contains three gromacs patches:
4.6.7: usual (old) hrex
5.0: no hrex
5.1: new hrex - to be tested well!

Notice that I just checked that it compiles and that it gives correct distribution of energies in an old test set that I have for alanine dipeptide. I used one CPU per replica, no openmp, and of course is a very specific case.

Let me know if you find problems!

Davide: I did not have time to include the changes you guys made for RANDOM_EXCHANGES. I will try to have a look at it and perhaps introduce it in the new version directly. As a side comment, notice that gromacs demux.pl tool have issues with RANDOM_EXCHANGES (look for an earlier post by Richard Cunha on this mailing list).

Ciao!

Giovanni



Chris Neale

unread,
Mar 30, 2016, 12:34:07 PM3/30/16
to PLUMED users, sebastian....@googlemail.com
amazing! Thank you very much Giovanni. I will report back with testing results. I will compare to the testing that I already did for the 4.6.7 version.

Again, I am really happy that you are so willing to share these things.
Chris.

nicolas...@gmail.com

unread,
Apr 1, 2016, 8:55:32 AM4/1/16
to PLUMED users, sebastian....@googlemail.com
Thank you very much Giovanni. 
* You mention a new hrex: can you please give more details about what is different with respect to the version in Gromacs4.6.7 ?
* Is the new hrex working for Gromacs5.1, 5.1.1 and 5.1.2, or only one of them?

Thanks again.

Nicolas

Giovanni Bussi

unread,
Apr 1, 2016, 9:01:40 AM4/1/16
to plumed...@googlegroups.com, Sebastian Schneider
Hi,

it is a port of the old implementation for gmx 4.6 to gmx 5.1. It is supposed to work exactly in the same way (if there are no bugs).

Concerning the specific version, I did the patch for gmx 5.1.2 (latest version), so I think it is safer to use that one.

Giovanni

To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/5c0659be-532d-4e27-ba50-a92f526eae40%40googlegroups.com.

Chris Neale

unread,
Apr 1, 2016, 4:51:57 PM4/1/16
to PLUMED users

Test system #1: box with two chloride ions in vacuum

(1a) run standard gromacs gromacs5.1.2 replica exchage with two replicas (one at 300 K and the other at 300.01 K) -- this seemed to work as expected.

(1b) run gromacs5.1.2 patched with plumed2.2.2-hrex with two replicas, both at 300 K and both with lambda=0 -- as expected, all energies are zero and all exchange attempts have P=1.0

(1c) as #1b except with both lambdas=1 -- as expected, all exchange attempts have P=1.0

(1d) as #1b except with both lambdas=0.5 -- as expected, all exchange attempts have P=1.0

(1e) as #1b except with lambdas at 1.0 and 0.001 -- as expected, exchange attempts have P=0.0. I then use gmx traj to get the full precision coordinates (saved using xtc-precision=1000000) and evaluate the energies, which match incredibly well (this is a double precision run):


me: 81.5086 plumed: 8.151e+01 from distances: 0.914412 and 0.394249 at time 0.002

me: 80.4921 plumed: 8.049e+01 from distances: 0.912717 and 0.396451 at time 0.006

me: 79.4887 plumed: 7.949e+01 from distances: 0.911223 and 0.398699 at time 0.01

me: 78.4993 plumed: 7.850e+01 from distances: 0.909893 and 0.400985 at time 0.014

me: 77.5297 plumed: 7.753e+01 from distances: 0.908568 and 0.403266 at time 0.018

…

me: 56.8677 plumed: 5.687e+01 from distances: 0.9371 and 0.478046 at time 0.122

me: 56.3657 plumed: 5.637e+01 from distances: 0.940528 and 0.481002 at time 0.126

me: 55.9071 plumed: 5.591e+01 from distances: 0.944196 and 0.483872 at time 0.13

…


However, when one of the replicas experiences particle separation that go outside the cutoffs, plumed doesn't seem to understand that. Below, you'll see that at time >=0.186 ps the energies diverge.


…

me: 51.5971 plumed: 5.160e+01 from distances: 0.983643 and 0.513695 at time 0.17

me: 51.2172 plumed: 5.122e+01 from distances: 0.988269 and 0.51678 at time 0.174

me: 50.8628 plumed: 5.086e+01 from distances: 0.993142 and 0.519835 at time 0.178

me: 50.5459 plumed: 5.055e+01 from distances: 0.998129 and 0.52276 at time 0.182

me: 105.687 plumed: 5.005e+01 from distances: 1.00308 and 0.525746 at time 0.186

me: 105.067 plumed: 4.943e+01 from distances: 1.00781 and 0.528857 at time 0.19

me: 104.428 plumed: 4.879e+01 from distances: 1.01259 and 0.532105 at time 0.194

me: 103.777 plumed: 4.814e+01 from distances: 1.01738 and 0.535455 at time 0.198


It gets stranger in that when I re-evaluate the energies assuming that the cutoff was much larger (same simulation data, just not truncating at 1 nm), then I get this:


…

me: 50.5459 plumed: 5.055e+01 from distances: 0.998129 and 0.52276 at time 0.182

me: 50.2188 plumed: 5.005e+01 from distances: 1.00308 and 0.525746 at time 0.186

me: 49.859 plumed: 4.943e+01 from distances: 1.00781 and 0.528857 at time 0.19

me: 49.4801 plumed: 4.879e+01 from distances: 1.01259 and 0.532105 at time 0.194

me: 49.0873 plumed: 4.814e+01 from distances: 1.01738 and 0.535455 at time 0.198


And you will note that the match is not really so good after the point where the cutoff was crossed, though the approximate value is now correct. 


One can fix this entirely by adding the following lines to the .mdp file:

coulomb-modifier=none

vdw-modifier=none


e.g. (from a different simulation since the simulation had to be conducted with new .mdp options):

...

me: 192.512 plumed: 1.925e+02 from distances: 1.25717 and 0.32504 at time 0.386

me: 191.909 plumed: 1.919e+02 from distances: 1.26005 and 0.325476 at time 0.39

me: 191.256 plumed: 1.913e+02 from distances: 1.26297 and 0.325955 at time 0.394

me: 190.491 plumed: 1.905e+02 from distances: 1.26591 and 0.326524 at time 0.398


I'm not sure exactly what's going on here. Anyway, I think there is some issue when using the default Potential-shift-Verlet modifiers.


I'm going to continue testing with some more systems that are a bit more realistic and will report back, but if anybody can make any sense of the fact that problems arise with coulomb-modifier=Potential-shift-Verlet and vdw-modifier=Potential-shift-Verlet (default options for gromacs 5.1.2), then I am very interested to hear about it.


Mdp file:

cutoff-scheme = verlet

constraints = all-bonds

lincs-iter = 1

lincs-order = 6

constraint_algorithm = lincs

integrator = sd

dt = 0.002

tinit = 0

nsteps = 200

nstcomm = 1

nstxout = 1

nstvout = 1000000000

nstfout = 1000000000

nstxtcout = 1

nstenergy = 1

nstlist = 1

nstlog=0 ; reduce log file size

ns_type = grid

rlist = 1.0

rvdw = 1.0

rcoulomb = 1.0

coulombtype = cut-off

tc_grps = System

tau_t = 1.0

ld_seed = -1

ref_t = 300

gen_temp = 300

gen_vel = yes

unconstrained_start = no

gen_seed = -1

Pcoupl = no

nstcalcenergy=1

xtc-precision=1000000

coulomb-modifier=none    ;; only included for last test

vdw-modifier=none    ;; only included for last test


####


Thank you,

Chris.

Chris Neale

unread,
Apr 1, 2016, 7:16:51 PM4/1/16
to PLUMED users
My mistake. The energy differences for exchanges are actually correct when one group is beyond the cutoff. In my evaluations of the energies, I forgot to subtract the energy at the cutoff from all values when using coulomb-modifier=Potential-shift-Verlet and vdw-modifier=Potential-shift-Verlet.

...
me: 89.8444 plumed: 8.984e+01 from distances: 0.98011 and 0.384475 at time 0.094
me: 90.3134 plumed: 9.031e+01 from distances: 0.983935 and 0.383935 at time 0.098
me: 90.7608 plumed: 9.076e+01 from distances: 0.987846 and 0.383452 at time 0.102
me: 91.2605 plumed: 9.126e+01 from distances: 0.991835 and 0.382865 at time 0.106
me: 91.8254 plumed: 9.183e+01 from distances: 0.996087 and 0.382171 at time 0.11
me: 92.3573 plumed: 9.236e+01 from distances: 1.00049 and 0.381508 at time 0.114
me: 92.7267 plumed: 9.273e+01 from distances: 1.00492 and 0.380733 at time 0.118
me: 93.1025 plumed: 9.310e+01 from distances: 1.00923 and 0.379952 at time 0.122
me: 93.4345 plumed: 9.344e+01 from distances: 1.01353 and 0.379268 at time 0.126
me: 93.7742 plumed: 9.377e+01 from distances: 1.01775 and 0.378574 at time 0.13
me: 94.1445 plumed: 9.414e+01 from distances: 1.022 and 0.377824 at time 0.134
...

Therefore, there are no issues so far with any of my testing.

Thank you,
Chris.

Chris Neale

unread,
Apr 6, 2016, 7:02:30 PM4/6/16
to PLUMED users
Dear Giovanni:

Looks like the box dimensions are not getting updated in hrex ( https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!topic/plumed-users/A6keTq-EWAw ).

There may be more problems with pressure coupling beyond that though I can't be sure yet.

Can you please direct me to where in the non-hrex code this evaluation of different energies is occurring and in comparison where you do it with the hrex code?

Thank you for your help,
Chris.

nicolas...@gmail.com

unread,
Jun 1, 2016, 10:35:54 AM6/1/16
to PLUMED users
Dear Giovanni,

I want to report that I think there is a bug somewhere in the Plumed2.2-hrex version. My system is protein+ligand in organic solvent. I am using 8 replicas from 300-800K, with also umbrella sampling.

With Plumed2.1/Gromacs4.6.7, here are the final statistics:

Replica exchange statistics

Repl  49999 attempts, 25000 odd, 24999 even

Repl  average probabilities:

Repl     0    1    2    3    4    5    6    7

Repl      .60  .62  .64  .65  .66  .67  .68

Repl  number of exchanges:

Repl     0    1    2    3    4    5    6    7

Repl     15023 15662 15943 16274 16519 16734 17047

Repl  average number of exchanges:

Repl     0    1    2    3    4    5    6    7

Repl      .60  .63  .64  .65  .66  .67  .68


Whereas with Plumed2.2/Gromacs5.1.2, here are the final statistics:

Replica exchange statistics

Repl  9999 attempts, 5000 odd, 4999 even

Repl  average probabilities:

Repl     0    1    2    3    4    5    6    7

Repl      1.0  .56  1.0  .58  1.0  .62  1.0

Repl  number of exchanges:

Repl     0    1    2    3    4    5    6    7

Repl     5000 2817 5000 2925 4999 3057 5000

Repl  average number of exchanges:

Repl     0    1    2    3    4    5    6    7

Repl      1.0  .56  1.0  .59  1.0  .61  1.0


I am not sure to know where to look for to give more information. Please let me know what I can do to help.


Nicolas

Giovanni Bussi

unread,
Jun 1, 2016, 10:53:44 AM6/1/16
to plumed...@googlegroups.com
Hi,

did you check that neighbor lists are updated with a stride that is an integer divisor of the stride between exchanges?

It looks like acceptance is wrong only in odd exchanges.

Giovanni


To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/3bd13b10-bfba-4165-b188-ce3635cc71d4%40googlegroups.com.

nicolas...@gmail.com

unread,
Jun 1, 2016, 11:07:20 AM6/1/16
to PLUMED users

Hi,


In both cases I have used the same input files (except the pull-code part that needs to be changed between Gromacs4.6.7 and Gromacs5.1.2). In particular, in the .mdp "nstlist = 20" and in the command line "-replex 100". So the answer to your question is yes.


Nicolas

Giovanni Bussi

unread,
Jun 1, 2016, 11:09:04 AM6/1/16
to plumed...@googlegroups.com
Are you sure that gromacs is not setting the update stride to 40? Can you retry using replex 200 to see if this solves the issue?

Giovanni


To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/1da7194b-6dcb-4666-9118-9633f210dfe0%40googlegroups.com.

nicolas...@gmail.com

unread,
Jun 1, 2016, 11:11:29 AM6/1/16
to PLUMED users
You're right, I have just seen this in one of the log file: "Changing nstlist from 20 to 40, rlist from 1.2 to 1.2"

I will change replex and keep you posted. Thanks a lot for the quick (and efficient) answer.

Nicolas

nicolas...@gmail.com

unread,
Jun 3, 2016, 3:35:56 AM6/3/16
to PLUMED users, nicolas...@gmail.com
Hi,

Just to confirm that it was indeed the problem. Statistics are now (Plumed2.2/Gromacs5.1.2):

Replica exchange statistics

Repl  4999 attempts, 2500 odd, 2499 even

Repl  average probabilities:

Repl     0    1    2    3    4    5    6    7

Repl      .60  .60  .62  .63  .65  .66  .67

Repl  number of exchanges:

Repl     0    1    2    3    4    5    6    7

Repl     1534 1467 1591 1531 1634 1683 1697

Repl  average number of exchanges:

Repl     0    1    2    3    4    5    6    7

Repl      .61  .59  .64  .61  .65  .67  .68



I did a small benchmark. Everything is the same (system, starting points, input files, etc...). I am using 8 replicas with HREX+Umbrella Sampling (the pull code has changed between the two versions of Gromacs, which is very likely the reason for the change in performances). Nodes have 12 cores, I used either 8 cores in one node or used 2 nodes (24 cores)

                                                            ns/day       hours/ns

Gromacs4.6.7/Plumed2.1, 8cores       2.462        9.749

Gromacs4.6.7/Plumed2.1, 24cores     5.180        4.633

Gromacs5.1.2/Plumed2.2, 8cores        3.165        7.582

Gromacs5.1.2/Plumed2.2, 24cores      6.763        3.549


Nicolas

rajeswar...@gmail.com

unread,
May 18, 2018, 5:24:30 AM5/18/18
to PLUMED users
Dear Giovanni,
I am using plumed 2.3.5 version and doing REST2 simulation. This is just to make sure whether have you incorporated the RANDOM_EXCHANGES with -hrex?

Thank you,
Rajeswari

Giovanni Bussi

unread,
May 18, 2018, 8:08:35 AM5/18/18
to plumed...@googlegroups.com
Hi,

no I think this is still open:


Giovanni


To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/5c66e896-4247-4505-9746-59c9a9d3dccd%40googlegroups.com.

Rajeswari A.

unread,
May 18, 2018, 9:21:41 AM5/18/18
to plumed...@googlegroups.com
Thanks for your reply. But when I'm running hrex with random exchanges, gromacs or plumed didn't complaint and runs properly (I mean, the exchanges truely happening randomly). Is there any mistake hidden while using together hrex and random exchanges?

Please advise me. 

Thanks once again for your quick reply,
Rajeswari

You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/eJ0xpnHPb_s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users...@googlegroups.com.

To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.

Giovanni Bussi

unread,
May 18, 2018, 10:05:22 AM5/18/18
to plumed...@googlegroups.com
Acceptance won’t be calculated correctly.

When you use hrex you should always do the recommended test (identical topologies should lead to acceptance=1)


For more options, visit https://groups.google.com/d/optout.
--
Sent from Gmail mobile

Rajeswari A.

unread,
May 18, 2018, 11:26:57 PM5/18/18
to plumed...@googlegroups.com
Dear Giovanni,

I will do the suggested test and get back to you. 

Thanks,
Rajeswari


Rajeswari A.

unread,
May 21, 2018, 1:44:05 AM5/21/18
to plumed...@googlegroups.com
Dear Giovanni,

As you said, the exchange acceptance is not calculated correctly when using hrex and random exchanges together.

My problem is when I use hrex with standard neighbour exchange, I did not see proper diffusion of replica over the state space, though the acceptance probability is ~25%. To overcome this, I have tried a lot with varying the highest temperature, bin size between consecutive two replicas etc. But no luck in improving the diffusion. Hence I am desperately looking forward for incorporating the random exchange condition along with hrex. Will you be able to help me by providing the patch that Davide has given you? I can contribute to the community by testing and reporting of that.

Thanks a lot for your help.
Rajeswari

Rajeswari A,
SERB National Postdoctoral Fellow
Molecular Biophysics,
Indian Institute of Science,
Bangalore, India - 560 012

To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/eJ0xpnHPb_s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users+unsubscribe@googlegroups.com.

To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users+unsubscribe@googlegroups.com.
--
Sent from Gmail mobile

--
You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/eJ0xpnHPb_s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users+unsubscribe@googlegroups.com.

Giovanni Bussi

unread,
May 23, 2018, 9:14:15 AM5/23/18
to plumed...@googlegroups.com
Hi,

if you have replicas sorted by increasing lambda, using random exchanges would not actually increase the diffusivity: in practice, acceptance will always be zero for non-neighboring replicas. What you see is an indication that there is some other degree of freedom not well sampled even at low lambda. By analyzing the structures you might be able to figure out which is this hidden barrier and you might try to combine -hrex with some other method, or just find which other parts of the Hamiltonian you should scale

Giovanni


--
You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/eJ0xpnHPb_s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users...@googlegroups.com.

To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.
--
Sent from Gmail mobile

--
You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/eJ0xpnHPb_s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users...@googlegroups.com.

To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/CAOzJmL9VizCVS-wX%3DLR%2BRddqXGi2xVQmDi%2BtHe5Azx8vmqa%3DAA%40mail.gmail.com.

Rajeswari A.

unread,
May 24, 2018, 3:06:20 AM5/24/18
to plumed...@googlegroups.com
Dear Giovanni,

Thank you very much for your valuable guidance.



Rajeswari A,
SERB National Postdoctoral Fellow
Molecular Biophysics,
Indian Institute of Science,
Bangalore, India - 560 012

To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/eJ0xpnHPb_s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users+unsubscribe@googlegroups.com.

To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users+unsubscribe@googlegroups.com.
--
Sent from Gmail mobile

--
You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/eJ0xpnHPb_s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users+unsubscribe@googlegroups.com.

To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.

--
You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/eJ0xpnHPb_s/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users+unsubscribe@googlegroups.com.

To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.
Reply all
Reply to author
Forward
0 new messages