[gmx-users] Using thermostats to create temperature gradient in system

[Agnivo Gosai]

Hello all,

Thanks to Dr. Lemkul for asking.

This is what I did to reach my present state :

(a) Prepared the topology of the system using GROMOS96 43a2 ff and energy
minimized the system after solvation using steepest descent algorithm.

(b) Did a position restrained ( protein restrained ) 5 ns NVT equilibration
using V-rescale thermostat at target temp. of 300 K.

(c) Then I kept the solvent at 300 K in a NVT ensemble and heated the
protein only by simulated annealing over 500 ps , to 400K. I checked the
individual temperatures and found that it was achieved.

(d) Then as mentioned in the paper if I remove the thermostat on the
protein then "grompp" gives the following error :-

Fatal error:
2936 atoms are not part of any of the T-Coupling groups
For more information and tips for troubleshooting, please check the GROMACS

I want to let my protein loose heat and thus go down a temp. gradient while
the solvent i.e. water stays at 300 K. However , I am unable to do this. If
I do not specify any thermostat in a NVT ensemble it becomes NVE and I only
get a single temperature profile.

Any hints as to how I can achieve my objective ?

Sorry for the previous incomplete post.


Thanks & Regards
Agnivo Gosai
Grad Student, Iowa State University.