This subproject is one of many research subprojects utilizing theresources provided by a Center grant funded by NIH/NCRR. The subproject andinvestigator (PI) may have received primary funding from another NIH source,and thus could be represented in other CRISP entries. The institution listed isfor the Center, which is not necessarily the institution for the investigator.ABSTRACT: Approaches for improving reconstruction quality will include refinement of alignment procedures with and without fiducial markers, and optimization of reconstruction methods, making use of a priori knowledge where appropriate. Extension and improvement of our recently developed algorithm for molecular motif searching will be an important component of this aim. The visualization bottleneck represented by 3D surface rendering of tomographic reconstructions will be addressed by developing automated model-based segmentation methods. Achieving the ambitious goal of mapping the macromolecular interactions within the cell will require reconstructions with high resolution (perhaps 2 nm) and algorithms that can detect and orient structural motifs contained within 3-D density maps having low signal-to-noise ratio and reconstruction artifacts (due to anisotropic data). The three approaches we are taking are:1. New approaches for simultaneous alignment and reconstruction that should greatly improve resolution in tomographic density maps. These approaches, which do not rely on fiducial markers, will benefit all tomography projects requiring high resolution, as well as those in which the specimens do not have such markers, or in which there is movement of the specimen relative to the markers (i.e., poorly attached frozen-hydrated sections, see below). 2. Extending and improving the existing program for motif searching (RAMOS; Rath et al., 2004). This development will benefit a large subset of current and future projects involving the in situ organization of macromolecular assemblies. The primary system will be muscle triad junctions (Wagenknecht) whose membrane surfaces contain numerous copies of the distinctive ryanodine receptor. First applied to isolated triad junction vesicles, the approach would be extended to triad junctions in situ with successful development of methodologies for sectioning or FIB-milling frozen-hydrated tissue 3. A method for automated segmentation of membranes in tomographic reconstructions. The manual tracing of membrane contours in thin slices from 3-D density maps is unsatisfactory for many reasons. First and foremost, it is highly subjective. Tracings by different individuals in the same laboratory can vary significantly, and could be influenced by results one expects to observe. Also, the process of membrane tracing is a major bottleneck in projects in which numerous density maps must be analyzed, e.g. studies of muscle mitochondria, each of which can contain scores of densely packed cristae. Automating the process of membrane segmentation would benefit several collaborative and service projects.
Showing the most recent 10 out of 252 publications
