Re: [AMBER] Restraining distance between a protein atom and a water of crystalization

From: Carlos Simmerling <carlos.simmerling.gmail.com>
Date: Fri, 22 Apr 2011 07:20:39 -0400

You should not use such large force constants. They give high frequency
motions, ans in then require short time steps to properly integrate.

Why not use a value of 10-100 and set larger r4? You can always use the
harmonic oscillator energy equation to estimate the amount of thermal
distortion you will get.

On Apr 22, 2011 1:13 AM, "nicholus bhattacharjee" <
nicholusbhattacharjee.gmail.com> wrote:

Dear Carlos,
                   Thank you for your help. I have gone through the manual
and could understand the restraint energy functions. But I am facing some
problems. First of all I have changed to ialtd=1 instead of ialtd=0 since it
gives a stiffer value (hyperbolic rather than linear) beyond r4. Now as far
I could understand We can restrain the distance by either monitoring r1->r4
values or maximizing rk2 and rk3 values. Since my restrained value is very
low of around 1.8 A there is low probability of the distance coming to lower
value during MD. So if I keep the rst.dist files like bellow
&rst
 ixpk= 0, nxpk= 0, iat= 1735, 2946, r1= 0.1, r2= 0.3, r3= 1.8, r4= 2.3,
     rk2=0.0, rk3=2000.0, ir6=1, ialtd=1,
 &end

By either giving high rk3 value (as shown above) or by giving high r4 value
I can keep the distance constrained in the desired value of around 1.8A.

This was fine with the minimization protocol (both minimization constraining
the solute as well as the whole system). But I am unable to run the
equilibrium run with this. During equilibration (where I am inceresing the
temperature from 100K to 300K) we are unable to give very high rk3 value as
it is showing problem with SHAKE where vlimit exceed. Neither giving very
high value to r4 provides a desired result i.e. I am unable to restrain the
desired distance. I am confused what should I do next. Kindly suggest some
solution.

On Mon, Apr 18, 2011 at 9:33 AM, Carlos Simmerling <

carlos.simmerling.gmail.com> wrote:

> I don't think you should use the script- what you were doing ...
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Apr 22 2011 - 04:30:03 PDT
Custom Search