[gmx-users] proper use of restraints

[paul buscemi]

Dear Users,

I’ve modeled a simple polymer ( nylon ), 600 atoms aligned in the x direction  .   I create a restraint file for one molecule using the top file indices and fx,fy fz = 10000 0 0  then place 10 to 100 copies in a box  On equilibration, they  exhibit the expected hydrogen bonding using the gromos54a7 ff.   Under npt pcouple - surface tension, the molecules remain elongated  and some coalesce  into groups but do not fully group  with ref p = 1 1  and comp = 4e-5  0 .  Just why the y direction does not shrink is a puzzle but not the main concern.   The x dimension fo the box remains constant too but I assumed that is because the ends of the polymers are fixed,  But there is room for packing in the y dimension.

Under p couple = isotropic, the ends re-enter the opoosite  (x) sides as the box collapses in x, y x directions  ( the original xyz  300 300 40 A.   I was under the impression that under relatively strong x restraint, the x direction would not collapse because the x direction is fixed at the ends  ???  Or is the way to look at it, the end ARE fixed, but the world around them is shrinking ?

Any comments ( well most any good ones ) would be appreciate.


Reference is made to the restraint file both in the molecule itp, and the base top file as an include statement - which is probably redundant if not outright wrong.

Paul