Volumetric valencies

The molecular visualization tool devised by Preston MacDougall and Christopher Henze.

Our desire to see what the building-blocks of nature look like seems to be irresistible, even when it is meaningless to talk of their visual appearance in any normal sense. Of all the modern genres of representing the unseeable, none has offered a more beguiling parade of visual delights than the modelling of big molecules. From the polished glyptic formula kits of the nineteenth century to today's proliferating programs of computer modelling, the images have not only satisfied our cognitive urges but have also played key roles in understanding and predicting the properties of molecules that are operating at the very heart of life, disease and death.

As the now quaint-looking ball-and-stick constructions have been superseded by a range of computer programs, such as Per Kraulis's widely used MolScript, so it has become apparent that different modes of representation may highlight quite different kinds of structural information. Accordingly, each system of modelling can handle potentially different aspects of the properties of the molecule. The as-yet-unwritten history of the iconography of molecules would tell of complex symbioses between new kinds of information, new and old representational means, and the research questions that shape the visual grammar of the imagery.

MacDougall and Henze's image of the vitamin B12 complex (cyanocobalamin form) is the joint handiwork of classical and quantum agents.

One of the most beguiling of the new systems is that recently devised by the chemist Preston MacDougall, of Middle Tennessee State University, and Christopher Henze, a visualization specialist at the NASA Ames Research Center in California. Their images are based on the topological analysis of the distribution of electron probability density, the cornerstone of Richard Bader's quantum theory of atoms in molecules. Using the laplacian of the probability distribution, their models effectively plot the lumps and hollows, the extrusions and holes in the electron cloud around each nucleus. The resulting sculptures could be described as the joint handiwork of classical and quantum agents.

A fourth dimension is provided by the colour coding, in which a rainbow colour scale is taken as corresponding to a range of charge concentration, from 'cool' blue depletions to 'hot' red concentrations. The transition from green to yellow corresponds to the inflection between depletion and concentration, and white denotes the very highest concentrations. Even with this amount of modelled and coloured information, the static image of an elaborate molecule remains frustratingly complex to unravel. The process of visualization is completed by an animated fly-around facility and by an opacity-function editor that allows us to focus and re-focus interactively on features of special concern (see Theor. Chem. Acc. 105, 345; 2001).

The tangibly seductive nature of the tool is not in doubt. But is all the ingenuity worth the effort in scientific terms? The answer is a definite yes, in that it represents an extension of the power of three-dimensional models to predict chemical bonding. Not only do the lumpy configurations reveal very clearly the strong covalent bonding that provides the structural integrity of a molecule, but they are also very effective in denoting sites of noncovalent interactions, the weaker links that typically involve hydrogen atoms at the extremities of large molecules. The plastic modelling tool discloses topologies that permit the identification of reactive sites, including novel ones that are outside the purview of rule-based algorithms. The identification of such sites promises rich potential, most notably in drug design.

As MacDougall himself recognizes, the historical antecedents of such techniques of sculptural 'fitting' go back at least to the 'lock-and-key' model of substrate binding advocated by Emil Fischer at the end of the nineteenth century. There is even an echo of the ideas of Nicholas Lemery, author of Cours de chymie in 1675, who was once derided for his conjecture that “chemical combination between two substances, such as an acid and a base, might be accounted for by supposing that the particles of one were sharp,and those of the other porous, and that chemical combination was effected by the fitting of the points into the holes”.

There is a nice sense that with the new tool, as is so often the case in scientific visualization, fundamental kinds of visual satisfaction and scientific functionality nourish each other in equal measure. For the animated fly-around, see http://www.nas.nasa.gov/~chenze/Preston/b12.mpg