Iterative stable alignment and clustering of 2D transmission electron microscope images - PubMed (original) (raw)

Iterative stable alignment and clustering of 2D transmission electron microscope images

Zhengfan Yang et al. Structure. 2012.

Abstract

Identification of homogeneous subsets of images in a macromolecular electron microscopy (EM) image data set is a critical step in single-particle analysis. The task is handled by iterative algorithms, whose performance is compromised by the compounded limitations of image alignment and K-means clustering. Here we describe an approach, iterative stable alignment and clustering (ISAC) that, relying on a new clustering method and on the concepts of stability and reproducibility, can extract validated, homogeneous subsets of images. ISAC requires only a small number of simple parameters and, with minimal human intervention, can eliminate bias from two-dimensional image clustering and maximize the quality of group averages that can be used for ab initio three-dimensional structural determination and analysis of macromolecular conformational variability. Repeated testing of the stability and reproducibility of a solution within ISAC eliminates heterogeneous or incorrect classes and introduces critical validation to the process of EM image clustering.

Copyright © 2012 Elsevier Ltd. All rights reserved.

PubMed Disclaimer

Figures

Figure 1

Figure 1. Flowchart of Stable Alignment and Clustering

Only clusters comprising images with alignment parameters that are stable (at a given pixel error threshold) across several independent rounds of within-cluster alignment are retained. Images in unstable clusters are sent back to the unassigned image pool for reclustering. See also Figure S5.

Figure 2

Figure 2. Flowchart of Iterative Stable Alignment and Clustering

The membership of clusters generated by four semi-independent SAC runs is compared, and only clusters with reproducible membership are retained. Images in clusters with low reproducibility are sent back to the unassigned image pool for reclustering.

Figure 3

Figure 3. ISAC Results Obtained for the Data Set of EF-Tu Ribosomal Complex

(A) Raw EM images. (B) Selection of ISAC cluster averages matched to projections of the X-ray structure. (C) Common lines volume compared with a map derived from the X-ray model. Scale bar corresponds to 10 nm. See also Figures S1–S3 and S6–S11.

Figure 4

Figure 4. ISAC Analysis of hRNAPII EM Images

(A) Raw EM images of hRNAPII. (B) A selection of hRNAPII ISAC cluster averages matched to projections of the X-ray structure of yeast RNAPII (pdb 1WCM). (C) A 3D map of hRNAPII derived by applying common lines to the ISAC averages shown in (B) (left), compared to a map of the homologous yeast RNAPII (right) derived from its X-ray structure. Scale bars correspond to 10 nm in (A and B) and 5 nm in (C). See also Figure S4.

Figure 5

Figure 5. Conformational Variability of hRNAPII Revealed by ISAC Cluster Averages

(A) Selected hRNAPII ISAC averages showing changes in the position/ appearance of the clamp domain (marked by yellow arrowheads). (B) hRNAPII volumes obtained after CD-PCA analysis of resampled ISAC averages show variability in clamp structure. (C) Two 3D maps obtained by competitive refinement of hRNAPII images using the volumes in (B) as initial references show alternative conformations of the hRNAPII clamp domain. Scale bars correspond to 10 nm in (A) and 5 nm in (B and C).

Similar articles

Cited by

References

    1. Armache KJ, Mitterweger S, Meinhart A, Cramer P. Structures of complete RNA polymerase II and its subcomplex, Rpb4/7. J Biol Chem. 2005;280:7131–7134. - PubMed
    1. Baldwin PR, Penczek PA. The transform class in SPARX and EMAN2. J Struct Biol. 2007;157:250–261. - PubMed
    1. Burkard RE, Mauro D, Martello S. Assignment Problems. Philadelphia: Society for Industrial and Applied Mathematics; 2009.
    1. Duda RO, Hart PE, Stork DG. Pattern Classification. New York: Wiley; 2001.
    1. Grundel DA, Pardalos PM. Test problem generator for the multidimensional assignment problem. Comput Optim Appl. 2005;30:133–146.

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources