Although the methodology of molecular microscopy has enormous potential, it is time consuming and... more Although the methodology of molecular microscopy has enormous potential, it is time consuming and labor intensive. The techniques required to produce a three dimensional (3D) electron density map of a macromolecular structure normally require manual operation of the electron microscope by a skilled operator and manual supervision of the sometimes complex software needed for analysis and calculation of 3D maps. Typically it will take weeks to months to collect and analyze a dataset in order to reconstruct a map at 10-20Å resolution. It is generally agreed that in order to increase the resolution to a range where secondary structure will be discernable in the map (~7Å) will require an order of magnitude increase in the amount of data collected and analyzed. It is clear that this will only be practical as a mainstream technique if we can automate the imaging and analysis processes and greatly improve the overall throughput.
Proceedings IEEE International Symposium on Biomedical Imaging
Cryo-electron microscopy is widely viewed as a uniquely powerful method for the study of membrane... more Cryo-electron microscopy is widely viewed as a uniquely powerful method for the study of membrane proteins and large macromolecular complexes-subjects that are viewed as extremely challenging or impossible to study using x-ray or NMR methods. Although the methodology of molecular microscopy has enormous potential, it is time consuming and labor intensive. Our group has done extensive work to automate image acquisition and processing for cryo-EM. In this paper we will provide an overview of our automated system, called Leginon, and present results where we used tobacco mosaic virus (TMV) as a proof of concept.
We have reconstructed a three-dimensional map of keyhole limpet hemocyanin isoform 1 (KLH1), usin... more We have reconstructed a three-dimensional map of keyhole limpet hemocyanin isoform 1 (KLH1), using our automated data collection software, Leginon, integrated with particle selection algorithms, and the SPIDER reconstruction package. KLH1, a 7.9 MDa macromolecule, is an extracellular respiratory pigment composed of two asymmetric decamers, and presents an overall D 5 point-group symmetry. The reconstruction is in agreement with previous data published on molluscan hemocyanins. The reconstructed map (11.3 A A resolution, 3r criterion) was used to fit an available X-ray crystallography structure of Octopus dofleini Odg, solved at 2.3 A A [J. Mol. Biol. 278 (4) (1998) 855], with satisfactory results. The results validate the approach of automating the cryoEM process and demonstrate that the quality of the images acquired and the particles selected is comparable to those obtained using manual methods. Several problems remain to be solved however before these results can be generalized.
A new learning-based approach is presented for particle detection in cryo-electron micrographs us... more A new learning-based approach is presented for particle detection in cryo-electron micrographs using the Adaboost learning algorithm. The approach builds directly on the successful detectors developed for the domain of face detection. It is a discriminative algorithm which learns important features of the particle's appearance using a set of training examples of the particles and a set of images that do not contain particles. The algorithm is fast (10 seconds on a 1.3 GHz Pentium M processor), is generic, and is not limited to any particular shape or size of the particle to be detected. The method has been evaluated on a publicly available dataset of 82 cryo-EM images of keyhole lympet hemocyanin (KLH). From 998 automatically extracted particle images, the 3-D structure of KLH has been reconstructed at a resolution of 23.2Å which is the same resolution as obtained using particles manually selected by a trained user.
This paper presents our recent progress toward a fully automated system for cryo-electron microsc... more This paper presents our recent progress toward a fully automated system for cryo-electron microscopy that integrates instrument control, computer vision algorithms and machine learning techniques. It describes our image analysis strategies for detection and selection of filaments in highly noisy images using multi-level perceptual organization. At the signal level, we use the Canny edge detector to detect weak boundaries. Collinearity at the primitive level is employed to organize discontinuous edges into line segments with a complete description, by using the Hough transform followed by an algorithm to detect end points of line segments. At the structural level, line segments are grouped into filamentous structures by seeking parallelism and employing high-level knowledge. In addition, statistical methods are used to split two filaments if they are joined together end-to-end. The performance of the proposed approach has been tested and evaluated by applying it to high magnification images of tobacco mosaic virus.
Uploads
Papers by Yuanxin Zhu