I am drawing the SOLVENT isosurface around a protein, and used the MAP
keyword to colour using a cube file which consists of a grid over the
active site with values for the 'depth' from the protein 'surface'
(i.e. deep within the active site, the SOLVENT isosurface should be
blue, and closer to the surface, it should be red).
So far, so good. In fact, so far so very good - I didn't think it
would work, as I thought there would need to be an explicit 1:1
correspondance between the points on the SOLVENT isosurface and the
points in my cube file. I was very impressed by the result.
However, I am unsure as to how Jmol assigns the colors to the SOLVENT
isosurface for points which are outside the cube file. Looking at the
results, I assume it is doing some sort of interpolation (I should say
extrapolation really) based on the nearest grid points, despite these
grid points being very far away.
My preferred behaviour would be for Jmol to not show the SOLVENT
isosurface for points which are not within my cube file. What do you
think? I know that conceptually the MAP is only supposed to provide
colors, but I don't think it is sensible to allow Jmol to guess the
colours for points outside the grid. These are practically guaranteed
to be incorrect, and worse, misleading (as Jmol doesn't complain that
some isosurface points are outside the MAP).