From: Jérôme Hénin (jhenin_at_ifr88.cnrs-mrs.fr)
Date: Sun Mar 13 2011 - 14:45:43 CDT
Hi Francesco,
You seem to be restraining distances to a mean value of zero (the
"centers"), with a very high force constant. This could be the cause
of the instability. It is probably not what you wanted to do anyway.
Cheers,
Jerome
On 13 March 2011 13:49, Francesco Pietra <chiendarret_at_gmail.com> wrote:
> Hello:
> I am equilibrating a thoroughly minimized colvar-restrained protein in
> a membrane. Heating alone (ts=1fs, only TIP3 hydrogens rigid), without
> press equilibration, revealed problems with CLA-patched GLU. That is,
> what was a respectful HE2-CLA distance during the minimization,
> shortens too much below vdw and the procedure crashes. Of course I'll
> try the common remedies, in particular ts<1fs and even slower heating
> (it was starting 0.5K, frequency 25, reassign 5). However, having no
> experience about, I wonder whether the values in colvar can also need
> tuning for the case in point. What I used, just for the cause of the
> crash:
>
> **********************************
> colvar {
> name colvar9
>
> width 1.0
>
> lowerBoundary -10.0
> upperBoundary 10.0
>
> lowerWall -5.0
> upperWall 5.0
>
> lowerWallConstant 100.0
> upperWallconstant 100.0
>
> outputSystemForce yes
>
> distance {
> group1 {
> atomNumbers 20137 # CLA 421C(P3)
> }
> group2 {
> atomNumbers 17866 # OE2 of GLU277C(P3)
> }
> }
> }
>
>
> harmonic {
>
> name CLA_harmonic
>
> colvars { colvar1 colvar2 colvar3 colvar4 colvar5
> colvar6 colvar7 colvar8 colvar9 }
>
> centers 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
>
> forceConstant 100.0
> }
> ******************************
>
> This means that I have restrained OE2-CLA.
>
> No problems arose, in what I have done, with the other ligands of CLA
> (ARG, LYS).
>
> Thanks for any suggestion about the trend of values to use in colvar.
>
> francesco pietra
>
>
This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:19:55 CST