[gmx-users] Setup Box Size

[John Kerrigan]

Sigh!  The setup box size that I had listed in my tutorial only applies to
work with globular proteins and the range given there is probably too
small even for that work. The important focus here was keeping the box
dimension to greater than twice the length of the cutoff (this was my only
thought at
the time I wrote the tutorial over a year ago). Typically, we use 1.0 nm
as a minimum spacing for more rigid globular proteins
and 1.2 nm as a minimum for work with more flexible small polypeptides (as
reported by McCammon in the paper you reference).  We occasionally go to
even 
larger box sizes for things like small simple beta sheet
dimers, and simple helix models because these beasties tend to translate
out of smaller boxes.  That said, I see
no reason why you cannot use a smaller spacing as low as 0.8 nm if you are
only interested in studying the interior of a rigid globular protein (e.g.
an enzyme active site.).  Provided of course that your resulting box
dimension remains at > 2x(cutoff).  If you are interested in studying
the more flexible exterior loop regions of your globular protein, then you
may want to consider the larger spacing (> 1.0 nm).  

I will make a correction to the intro tutorial to cover this aspect.  It
seems to me that the spacing issue is in part dependent on the size and
flexibility of the system you are studying in addition to the cutoff and
preservation of water structure.  There may be other factors I'm missing.
These are important considerations and I thank you very much for bringing
this oversight in my tutorial to my attention.

Cheers,

John E. Kerrigan, Ph.D. 

Research Support, Academic Computing Services
Robert Wood Johnson Medical School
University of Medicine and Dentistry of NJ
675 Hoes Lane
Piscataway, NJ 08854   USA

(732) 235-4473  phone
(732) 235-5252  FAX