Molecular microscopy has become an increasingly important tool for structural biology but the methodology is very labor intensive and very slow. It is generally recognized that the development of improved capabilities for three-dimensional electron microscopy are critical for progress in emerging integrative research in molecular cell biology. We aim to develop a system for rapid routine structure determination of macromolecular assemblies. Our ultimate goal is to develop an integrated system that can produce a three-dimensional electron density map of a structure within a few hours of inserting a specimen in the electron microscope. The motivation for this work is to provide answers to interesting biological questions. We will initially use our work on motor-microtubule complexes and actomyosin as the driver for the development of the integrated system. By tightly coupling the development of the new system with its implementation in a laboratory whose primary goal is answering fundamental questions in cell biology, we will obtain immediate and invaluable feedback as to how the system is used in practice.

Developing this system will involve devising new approaches and integrating the results of several ongoing research projects. The primary specific aims are: (1) To remove the requirement for using film to acquire the high magnification electron micrographs. This will require the development of feature recognition algorithms and new imaging strategies that take into account the characteristics of currently available digital cameras. (2) Improve and automate our existing software for helical image analysis. We will incorporate new methods for determining the helical parametersof an unknown specimen, methods for improving the resolution, andmethods for analyzing non-helical specimens. (3) Integration of the acquisition and the analysis steps. This will require incorporation of machine learning techniques to produce a system that is highly efficient in terms of throughput and data quality.

The general framework for integrated acquisition and analysis to be developed will be readily extendible to other specimens (helical tubes, single particles, two-dimensional crystals). Thus, once the system has been successfully implemented it will be made generally available to the scientific community. The system we plan to develop has the potential to revolutionize the field of three-dimensional electron microscopy and make this approach accessible to a wide community.