Re: ABF - restricting reaction coordinate

From: Jerome Henin (jhenin_at_cmm.chem.upenn.edu)
Date: Fri Dec 07 2007 - 10:01:47 CST

Dear Vaithee,

The problem might come from the fact that the dihedral coordinate uses
force constants in kcal/mol/degree^2, which is a rather large unit. In
many cases, a force constant of 0.1 can be sufficient.

This does not explain why the coordinate gets out of range. Does it
start within the desired range? If not, it would be interesting to see
the initial behavior. Can you send me the full output (maybe
off-list)?

Best,
Jerome

On Dec 7, 2007 9:55 AM, Subramanian Vaitheeswaran <vaithee_at_umd.edu> wrote:
> Dear Chris and Jerome,
>
> Following both your suggestions a couple of days ago, I split the range of my reaction coordinate (a dihedral angle) into four windows. I specify the following in my .conf file:
> ********************************
> abf coordinate dihedral
> abf abf1 {8 6 12 16}
> abf dxi 1.0
> abf xiMin 0.5
> abf xiMax 45.5
> abf writeXiFreq 50000
> ********************************
> So, I expect that xi will be in the range [0.5, 45.5] degrees for *every* step.
>
> But xi is almost never in this range - here is a small fragment grepped from the log file:
> ********************************
> TCL: ABF> Xi at timestep 4300000 : -119.747074603
> TCL: ABF> Xi at timestep 4350000 : -44.6413954893
> TCL: ABF> Xi at timestep 4400000 : -55.1637937915
> TCL: ABF> Xi at timestep 4450000 : -108.865140314
> TCL: ABF> Xi at timestep 4500000 : -155.734544341
> TCL: ABF> Xi at timestep 4550000 : -57.9804339106
> TCL: ABF> Xi at timestep 4600000 : -160.710021078
> ********************************
> Does this imply that the default value for forceConst of 10 kcal/mol/^2 is too small for this system? Do you see any problems with increasing it to an arbitrarily large value?
>
> Thanks for all your help,
> Vaithee
>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:47:23 CST