Today, crystallographer Harry Greenblatt kindly looked at the page I made yesterday on electron density maps (http://www.umass.edu/molvis/tests/jmol_edm/ ). He pointed out that the 2 Angstrom resolution example that I chose, 3hz7, has an unusually poor electron density (e.g. all those waters and outlying residues that have no density at 1 sigma). Checking at MolProbity showed this model to have an unusually high number of atomic clashes for its resolution. So it is a bad example. I will replace it when I have time. Meanwhile, I have put a note about this on the page.
Eric Martz, Ph.D.
Professor Emeritus, Microbiology, University of Massachusetts, Amherst