The proposed research project is aimed at the development and the dissemination of computational methods that will be used in electron microscopy for tomographic studies at the molecular level. Originally, we have developed and implemented the methods for studying force production in muscle. We can show that these methods are also useful for a variety of other biological systems, including ice-embedded samples, to characterize structural heterogeneity. Our approach is based on electron tomography to obtain 3D reconstructions, and multivariate statistical analysis and classification for processing and interpretation of the resulting volumetric data. This differs from traditional single particle analysis, which is primarily concerned with 3D reconstruction from projection images, in that classification is used to group macromolecular motifs according to structural differences or conformations, instead of particle orientation. The further development and dissemination of the methodology will require effort in several areas. In order to obtain higher resolution, we need to create efficient procedures for contrast transfer function correction of tilt series data, and add the capability of processing dual axis tilt series to the software in order to produce tomograms with more isotropic resolution. In this context, effective algorithms for the treatment of the missing wedge, or missing pyramid for dual axis tilt series, will be implemented in the volume data analysis procedures. The methods will be validated with model data and tested with several types of specimens that are of great biological and clinical interest, including muscle sections, two-dimensional protein arrays on lipid monolayers, integral membrane proteins, and envelope proteins on SIV and HIV-1 virions. The software implementation of our procedures will be designed to be interoperable with other major software packages used in electron microscopy and will be distributed on the internet. This will ensure that our methods are readily available in other areas of biological research. ? ? ?
Showing the most recent 10 out of 11 publications
