Present state of researchSupramolecular machines perform their work in the cell by going through many different states, distinguished by different conformations and free-energy levels. Ideally, in order to find out how these machines work, we would create a suitable in vitro environment containing all components including energy supply that allows the machine to function. We would then aim to take a "movie," capturing their structure at highest resolution in a continuous fashion. Keeping within that film analogy, we might consider taking a large number of "snapshots" in equal small time intervals, each short enough, as in the macroscopic world, to eliminate jarring transitions. However, we would find out that this project has flaws both on the conceptual and the practical level. Conceptually, it is incorrect to equate a molecular machine's progress to the workings of a macroscopic machine in motion since the states are not ordered in sequence of time but are visited in a stochastic manner, with occasional irreversible events such as NTP hydrolysis as the only mark of progress. In practical terms, there are in fact two problems, one affecting the way data for any given state can be captured, the other affecting the ability to obtain coverage of states in a continuum.First of all, the visualization of a structure requires some form of radiation which imparts energy on the molecule and changes it in the very process. Minimization of these damaging interactions calls for a radiation dose so low that averaging over a sufficiently large population is required for visualization. This requirement, in turn, translates into a complicated way each snapshot must be taken: the structural information in every state has to be gathered from many different copies of the molecule. While in forming such an average, crystallographic approaches -X-ray and electron crystallography -are able to take advantage of the regular arrangement of molecules in a crystal, single-particle electron microscopy must first determine the precise orientation of each molecule from its projection image [1]. The second hurdle interfering with the idea of making a movie is that only a limited number of states are sufficiently populated to allow a three-dimensional structure to be determined. Thus the "movie frames" in between these states remain empty, undetermined.New Chemistry and New Opportunities from the Expanding Protein Universe Downloaded from www.worldscientific.com by RUTGERS UNIVERSITY on 08/17/15. For personal use only.