The proposed research aims to extend our capability to analyze the structure of biological macromolecules and their assemblies using high resolution electron microscopy in combination with image analysis. The long term goal is to achieve near atomic resolution for helical assemblies of macromolecules. Of importance are the development of algorithms that can detect high resolution signal hidden by noise and algorithms that permit automated data analysis so that large scale averaging can be used to extract the signal from the noise. The proposal seeks to combine the automated data analysis with automated data collection in order to obtain large data sets. By analyzing large data sets with the powerful methods of multivariate statistics, we can uncover those factors that govern the preservation of high resolution data in the specimens. Such an analysis provides one way to improve specimen preparation and microscopy. Equally important is the proposed development of electron diffraction for helical structures, which will supply Fourier amplitudes to accompany Fourier phases extracted from the images. The amplitudes and phases are both needed to produce accurate three dimensional maps of the helical structures. There are a host of helical structures of biological importance which require higher resolution study: for example, viruses, actin, microtubules, and the paired helical filaments found in Alzheimer's disease. The methods developed are to be tested and applied to actin, tobacco mosaic virus and bacterial flagella. Classical x-ray analysis does not have the same potential for structural studies of these helical assemblies. Electron microscopy has the potential not only for helical structures but also structures having lower symmetry. Some f the algorithms and procedures developed as part of this proposal will be applicable to such structures.
