Re: Fwd: [gmx-users] SD integrator

2013-09-15 Thread HANNIBAL LECTER
Thanks. From your publications it seems that you have been simulating with infinite CNTs. I am trying to use finite CNTs and I am generating my topology files according to http://chembytes.wikidot.com/grocnt. I am modeling the peripheral Carbon atoms with two bonds by specifying it in the

Re: Fwd: [gmx-users] SD integrator

2013-09-15 Thread Dr. Vitaly Chaban
Are you sure that the source of instability is the CNT, not the protein? If yes, what makes you to think so? If you freeze all the atoms of the CNT, does it work well? Dr. Vitaly V. Chaban On Sun, Sep 15, 2013 at 11:07 PM, HANNIBAL LECTER hanniballecte...@gmail.com wrote: Thanks. From your

Fwd: [gmx-users] SD integrator

2013-09-13 Thread Dr. Vitaly Chaban
There must be some problem with degrees of freedom in your system... Dr. Vitaly V. Chaban -- Forwarded message -- From: HANNIBAL LECTER hanniballecte...@gmail.com Date: Fri, Sep 13, 2013 at 7:15 PM Subject: [gmx-users] SD integrator To: Discussion list for GROMACS users

Re: Fwd: [gmx-users] SD integrator

2013-09-13 Thread Dr. Vitaly Chaban
Do you couple CNT as a separate T-group? It should be done so. I think your systems blows up because of incorrectly defined bonded parameters for the tube, not due to thermostatting. For instance, you might not define all the valence angles and dihedrals. See my papers for the parameters for