0%

​

Share this page with:

What are volume data?

Macromolecules inside living cells include DNA, RNA, proteins, lipids, and polysaccharides; as well as hybrid macromolecules such as the lipopolysaccharides found in the outer membrane of Gram-negative bacteria. These macromolecules can form short-lived complexes, such as kinases or phosphatases binding their substrate during a signalling cascade; or can form large assemblies such as the proteins and rRNA which make up the ribosome. Structural volume data describe the extent in three-dimensional space of these macromolecules and their complexes. The shape of macromolecules can provide us with important information on their function and interactions.

Volume data are generated by a number of experimental methods, including X-ray crystallography, electron microscopy, electron or soft X-ray tomography, and small angle X-ray or neutron scattering. In electron microscopy, volumes representing the electron potential of the sample molecule are reconstructed from many 2D projections extracted from micrographs. In crystallography, solution of the phase problem gives electron density maps; which are then invariably interpreted in terms of atomic models.

Figure 1  Examples of volume data from electron microscopy and tomography, taken from the Electron Microscopy Databank.

Volume data is typically represented in software or files as a set of density values on a rectangular grid, referred to as a map (see panel A below). That is, the continuous volume is sampled at regular intervals, with the sampling frequency typically higher than the intrinsic resolution of the data.

Figure 2 (A) Schematic of discretised volume data, with voxel intensities representing the density value at a grid point. (B) Mesh representation of ribosome volume data, as viewed in UCSF Chimera.

There are also a number of graphical representations of volume data, such as that shown in panel B above. Rather than show all density values held in a map file, graphical programs typically show a surface, derived from a suitable contour level of the density values.

Segmentation

Volume data can be segmented in three-dimensional space. Different segments may represent different component molecules in a complex, or may represent visually distinct regions in a tomogram. Segments are usually supposed to have clear boundaries, though in reality there may be some uncertainty as to where the border between two segments lies. There may also be a hierarchy of segments.