Search

Electron tomography (ET) has been proven an essential technique for imaging the structure of cells beyond the range of the light microscope down to the molecular level. Large-field high-resolution views of biological specimens span more than four orders of magnitude in spatial scale, and, as a consequence, are rather difficult to generate directly. Various techniques have been developed towards generating those views, from increasing the sensor array size to implementing serial sectioning and montaging. Datasets and reconstructions obtained by the latter techniques generate multiple three-dimensional (3D) reconstructions, that need to be combined together to provide all the multiscale information. In this work, we show how to implement montages within TxBR, a tomographic reconstruction software package. This work involves some new application of mathematical concepts related to volume preserving transformations and issues of gauge ambiguity, which are essential problems arising from the nature of the observation in an electron microscope. The purpose of TxBR is to handle those issues as generally as possible in order to correct for most distortions in the 3D reconstructions and allow for a seamless recombination of ET montages.

Electron tomography is a powerful tool for deriving three-dimensional (3D) structural information about biological systems within the spatial scale spanning 1 nm3 and 10 µm3. With this technique, it is possible to derive detailed models of subcellular components such as organelles and synaptic complexes and to resolve the 3D distribution of their protein constituents in situ. While there continues to be progress towards the integration of high-performance computing technologies with traditional electron tomography processes, there is a significant need for more transparent integration with applications and to minimize the administrative overhead and complexity (resource administration, authentication, scheduling, data delivery) passed on to the noncomputer scientist end user. Here we present the "Telescience Portal" (https://gridport.npaci.edu/Telescience) as an example of a fully integrated, web-based solution for performing end-to-end electron tomography. More than just a collection of individual applications, the Portal provides a transparent workflow, where simple intuitive interfaces for grid-enabled parallel computation, resource scheduling, remote instrumentation, advanced image processing and visualization, access to distributed/federated databases, and network-enabled data management and archival are tightly coupled within a secure environment which promotes increased collaboration between researchers. This tightly integrated Telescience system is a test-bed application for using grid resources to accelerate the throughput of data acquisition and processing, increase access to scarce and/or expensive instrumentation, and improve the accuracy of derived data products.