Protein graphics: historical development, future directions and microcomputer applications

Designers in the macroscopic world of highrise buildings, spacecraft and robots can see the relationship between structure and function in their designs. This is not possible for protein chemists. Yet the relationship between structure and function is just as critical in the unseen world of molecules as it is in the world that we can touch and see. Engineering graphics in computer-aided design and manufacturing has increased efficiency in conventional design and made possible design of complex systems such as industrial robots. The new design possibilities opened to engineers by interactive computer graphics are indeed revolutionary, but they still may be superseded by graphics in protein research and engineering. Although the capability to visualize simple molecular structures in and of itself often gives biologists important clues as to the function of proteins, computer graphics also has given biologists a tool that they can use to calculate structural properties of proteins. Predictions from the modeling studies can be tested experimentally, providing a continual interplay between theory and practice. Unfortunately, at present, only about 100 of the over 4000 proteins with known amino acid sequences can be displayed even though, in principle, the complete three-dimensional structure of a protein is determined by the minimum energy configuration of its amino acid sequence. Only 100 or so proteins have known atomic coordinates because currently the only means of determining these coordinates experimentally is by X-ray crystallography, an expensive method that usually takes about 5 years from start to finish. Even though the set of proteins with known atomic coordinates is growing and the time required to solve protein structures using X-ray crystallography is being reduced, a dramatic improvement in the understanding of protein structure and function requires a synergy of techniques that bring together all the information on protein structure and function. These techniques include interactive graphics, molecular mechanics, distance geometry, quantum mechanics and analysis of experimental observations using expert systems.