On Fri, Aug 20, 2010, William Flak wrote:
> * In AMBER 10 manual, starting coordinate files must be the same for the
> two states, perturbed and unperturbed. Accoriding to the input group
> file:
> -O -i mdin -p prmtop.0 -c eq1.x -o md1.o -r md1.x -inf mdinfo
> -O -i mdin -p prmtop.1 -c eq1.x -o md1b.o -r md1b.x -inf mdinfob
> So, should I equilibrate ONLY one state and use its coordinates to
> describe the both states?
Yes.
>
> * In the manual, The implementation requires that the masses of all
> atoms be the same on all threads. How!!! I see papers studied systems as
> methanol to methane. Here the masses are not the same.
In classical mechanics, masses only affect the kinetic energy, not the
potential energy. Since you invariably subtract free energies on two sides
of a cycle (that have the same "real" mass change), the masses used by the
program do not affect the final results. If I remember correctly, the program
uses the masses of the first group, so you don't have to worry about this.
If you like, do a second calculation swapping group 1 with group 2: the code
will now use the masses of group 2, but you should get the same results, to
within the statisical uncertainty of your simulation.
....hope this helps....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Aug 21 2010 - 06:30:04 PDT
