Electron Density Map Demonstration
in Jmol 11.9.20_dev with CCP4 map format
Eric Martz*, January 19, 2010.

This page is obsolete. There is nothing wrong with the electron density map for 3hz7. I later realized that the problem is with Jmol's display of the map, not with the map itself. Please see THIS UPDATE.

85% of the macromolecular structures available from the Protein Data Bank (PDB) were determined by X-ray crystallography. A crystallographic experiment produces an electron density map for the average unit cell of the protein crystal. The amino acid (or nucleotide) sequence of the crystallized polymer(s) is known in advance. The crystallographer fits the atoms of the known molecules into the electron density map, and refines the model and map to the limits of the resolution of the crystal (which is limited by the level of order or disorder in the crystal). The crystallographer then deposits a model of the asymmetric unit of the crystal in the PDB, along with the experimental diffraction data (amplitudes and widths of the X-ray reflection spots, or structure factors) from which the electron density map can be reconstructed.

Examining the correspondence between the model and the electron density map (EDM) provides much clearer insight into the uncertainties in the model than does merely examining the model itself (see also Quality assessment for molecular models). Crystallographers generally use heavy duty visualization and modeling software such as Coot or PyMOL, which require considerable practice to use effectively. Jmol first became capable of displaying electron density maps in January, 2010. Being able to display EDM's in Jmol opens the door to examining EDMs effectively in a web browser, with a user interface (yet to be developed) that requires no specialized software knowledge. Until such an interface becomes available, the methods demonstrated here can be utilized.

Electron density maps are available for most PDB files from the Uppsala Electron Density Map Server. Get the map file in CCP4 format. Files used here were renamed {PDB_code}_map.ccp4. You can then display it in Jmol using commands shown below.

3hyd was chosen because it is small (a peptide of 7 amino acids in beta-strand conformation), with good resolution (1.0 Å). Note that the model appears to have been refined with all hydrogen atoms included. (Click on each command below, in sequence. Drag on the molecule with your mouse to rotate it; drag up and down on the right edge of the Jmol rectangle to zoom in/out.)
The electron density map file contains the density for the entire unit cell. Now we'll trim the mesh to be within 2 Å of atoms in the asymmetric unit.
The same capability that enables the map to be trimmed to the asymmetric unit (all atoms, {*} in the above command) can be used to show the map for a single residue.
Notice that even at a resolution of 1.0 Å (quite unusual for protein crystals) hydrogen atoms have almost no discernable electron density. At this resolution, each carbon, oxygen, or nitrogen atom has a distinct spherical electron density. Below, we'll compare with a resolution of 2.0 Å.

The above maps are contoured at 1.0 sigma (cutoff 1.0 in the above Jmol commands). Since Jmol's ability to display EDMs is new, at left is shown a comparison of the renderings at 1.0 sigma for PyMOL vs. Jmol. It appears that Jmol draws the map at about 96% of the diameter drawn by PyMOL, a difference that seems unimportant.

Crystallographers fit atoms into maps that are contoured at 1.0 to 1.5 sigma, according to taste. The difference for most atoms is very small, as shown in this comparison (displays in Jmol above): 1.0 sigma vs. 1.5 sigma.

However, 1.5 sigma is more conservative when there is greater disorder, as can be seen with a sidechain carbon (centered) in the N-terminal Leu1. (Displays in Jmol below.)

It has come to my attention (thanks to crystallographer Harry Greenblatt) that 3hz7 is not a good example. For its resolution, its electron density is poor, there are too many atoms outside 1 sigma (including nearly all waters), and the difference map has too much density unaccounted for. Further investigation showed that it has a very high number of atomic clashes. For now, what follows should be taken as an example of a model of poor quality. When time permits, I will replace it with a better example. (January 20, 2010)

3hz7 is a small protein domain (87 amino acids) with a resolution of 2.0 Å, which is the median resolution for the PDB.
Note that some outlying residues have been modeled despite an absence of electron density reaching 1.0 sigma. Few of the hetero atoms fall within 1.0 sigma.
The maps examined above are full maps (2mFo - DFc). Crystallographers typically examine a difference density map (mFo - DFc) to identify  positive difference densities  not accounted for by atoms in the model, or negative difference densities which could mean that a nearby atom is in the wrong position, or that atom was disordered by radiation damage. Difference density maps are usually contoured at 3.0 sigma. Below, we see many positive densities at +3.0 sigma, but no negative densities at -3.0 sigma, and still none at -1.5 sigma. Only at -0.5 sigma a negative density appears around a single atom, the sulfur oxide on Cys11, SX.S. I HAVE NOT YET VERIFIED WITH TRUSTED SOFTWARE THAT JMOL IS HANDLING THE NEGATIVE DIFFERENCE DENSITIES CORRECTLY.
Finally, as an alternative to restricting the map to a subset of atoms, we can use slabbing to examine the map in a complicated structure. Notice that at 2.0 Å resolution we no longer see discrete spheres for each C, O, N. The density is "smeared".

Open index.html in your browser. It will work from downloaded files without changing anything.

*Thanks to crystallographers Michal Harel, Harry Greenblatt, Boris Brumfeld, and Joel Sussman for advice.