Automated data collection for electron microscopic tomography - PubMed (original) (raw)

Automated data collection for electron microscopic tomography

Shawn Q Zheng et al. Methods Enzymol. 2010.

Abstract

A fundamental challenge in electron microscopic tomography (EMT) has been to develop automated data collection strategies that are both efficient and robust. UCSF Tomography was developed to provide an inclusive solution from target finding, sequential EMT data collection, to real-time reconstruction for both single and dual axes. The predictive data collection method that is the cornerstone of UCSF Tomography assumes that the sample follows a simple geometric rotation. As a result, the image movement in the x, y, and z directions due to stage tilt can be dynamically predicted with the required accuracy (15nm in x-y position and 100nm in focus) rather than being measured with additional images. Lacking immediate feedback during cryo-EMT data collection can offset the efficiency and robustness reaped from the predictive data collection and this motivated the development of an integrated real-time reconstruction scheme. Moderate resolution reconstructions were achieved by performing weighted back-projection on a small cluster in parallel with the data collection. To facilitate dual-axis EMT data collection, a hierarchical scheme for target finding and relocation after specimen rotation was developed and integrated with the predictive data collection and real-time reconstruction, allowing full automation from target finding to data collection and to reconstruction of 3D volumes with little user intervention. For nonprofit use the software can be freely downloaded from http://www.msg.ucsf.edu/tomography.

Copyright © 2010 Elsevier Inc. All rights reserved.

PubMed Disclaimer

Similar articles

Cited by

Publication types

MeSH terms

LinkOut - more resources