AMBER Archive (2002)

Subject: Re: Periodic water box

From: Thomas Cheatham (cheatham_at_chpc.utah.edu)
Date: Wed Jun 19 2002 - 22:57:04 CDT

> Xleap. I have then minimized the proteins in water for 100000 steps. I

Note that this level of minimization, prior to dynamics, is probably
serious overkill. You can probably get by with simply 500-10000 steps and
save some time for dynamics.

> then let it equilbrate over a further 100000steps. As far as I know I
> have set up a periodic box , however looking at the structure after
> these 200000 steps it appears as if the water is in a diamond shape..i.e
> not adhering to the Periodic box at all. In the out put file EWALD seems

You hit on the key; it is a "periodic" box. Don't lose all your joy!

AMBER supports space-filling unit cells that are non-rectangular (i.e.
that have angles differing from 90 degrees). To put all the molecules
back into the periodic unit cell therefore requires, in general, more than
simply seeing if an atom has an x, y, or z coordinate greater than +/- 1/2
the box length. To do this, we transform the coordinates into reciprocal
space and then truncate values below zero or above 1 and then transform
back. If this is done for the truncated octahedron (which has angles of
109.47122...), the resulting shape is diamond-like rather than the more
common/familiar spherical shape.

My guess is that you used the solvateOct command in LEaP (specifying the
"iso" keyword) and that after the initial dyamics, with IWRAP=1, the
resulting coordinates popped into the triclinic or diamond-like shape.
Had you not turned on the IWRAP=1 option, you would instead have seen the
water seemingly float away from the protein. Both "views" of the periodic
unit cell are scary at first sight since they seem problem but both are
equivalent as long as they collapse to the same unit cell.

Using the imaging facilities in ptraj, you can convert your box to be more
like expected (i.e. have your proteins maximally surrounded by water)
using the image command with the familiar option... If you do not specify
the familiar keyword, triclinic imaging will be performed (leaving the
diamond like shape).

As an example script, if your protein is residues 1-7500, you would do:

ptraj prmtop << EOF
trajin restrt
trajout restrt.imaged
center :1-7500 mass origin
image origin center familiar
EOF

Good luck with your simulations and let me know if you need further
clarification.

Thomas E. Cheatham, III
Department of Medicinal Chemistry & Center for High Performance Computing
University of Utah INSCC 418 / BPRP 295A
30 South 2000 East, Room 201 155 South 1452 East
Salt Lake City, Utah 84112-5820 Salt Lake City, Utah 84112-0190
tec3_at_utah.edu http://www.chpc.utah.edu/~cheatham
phone: (801) 587-9652 FAX: (801) 585-9119