Electron Microscopy: 3D reconstruction from 2D projections


In the hierarchy of biological visualization, 3D Electron Microscopy (EM) bridges the gap between the object sizes studied by X-ray crystallography and light microscopy. Single-particle EM is routinely used to resolve the three-dimensional structure of large macromolecular assemblies.

One of the major challenges in single-particle EM is structural heterogeneity of the studied particles. Particles can adopt different conformations and can be found in assemblies with alternative quaternary-structure. Computational methods for image processing and three-dimensional structure determination play a crucial role in single-particle EM. Most commonly used computational approaches assume that the imaged particles have homogeneous shape and quaternary-structure. When this assumption is violated, the product of the three-dimensional reconstruction is one low resolution structure. My aim is to improve and redesign various computational stages of particle reconstruction, taking into account the heterogeneity of the data.

Reconstruction of Multiple Structural Conformations of Macromolecular Complexes

We devised a new computational method that reveals the existence of different conformational states from EM data of macromolecular complexes. It is able to automatically classify the experimental images into homogeneous subsets which produce structurally different models. The method achieves high accuracy on synthetic data sets and real data. In one test case, we successfully differentiated between the real experimental images of human translation initiation factor eIF3 and of eIF3 complexed with hepatitis C virus (HCV) IRES RNA. In another experiment, where we used experimental images of human RNA polymerase II, our method produced two models that show a substantial conformational flexibility of the protein complex.

A Method for the Alignment of Heterogeneous Macromolecules

We proposed a feature-based image alignment method (Shatsky et al. '09) for single-particle EM that is able to accommodate various similarity scoring functions while efficiently sampling the two-dimensional transformational space. We use this image alignment method to evaluate the performance of a scoring function that is based on the Mutual Information (MI) of two images rather than one that is based on the commonly used cross-correlation function. We show that alignment using MI for the scoring function has far less model-dependent bias than is found with cross-correlation based alignment. We also demonstrate that MI improves the alignment of some types of heterogeneous data, provided that the signal-to-noise ratio is relatively high. These results indicate, therefore, that use of MI as the scoring function is well suited for the alignment of class-averages computed from single-particle images. Our method is tested on data from three model structures and one real dataset.


Software:

Contributed to:

Publications:

  • D. Zhi, M. Shatsky, SE Brenner. 2010. Alignment-free local structural search by writhe decomposition. Bioinformatics. in press.

  • M. Shatsky, R.J. Hall, E. Nogales, J. Malik, S.E. Brenner. Automated multi-model reconstruction from single-particle electron microscopy data. J Struct Bio. 170:98-108, 2010.

  • BG Han, M. Dong, H. Liu, L. Camp, J. Geller, M. Singer, TC Hazen, M. Choi, HE Witkowsa, DA Ball, D. Typke, KH Downing, M. Shatsky, SE Brenner, JM Chandonia, MD Biggin, RM Glaeser. Survey of large protein complexes in D. vulgaris reveals unexpected structural diversity. Proceedings of the National Academy of Sciences of the United States of America 106:16580-16585. 2009.

  • M. Shatsky, R.J. Hall, S.E. Brenner, and R.M. Glaeser. A method for the alignment of heterogeneous macromolecules from electron microscopy. J Struct Bio. 166:67-78, 2009.

  • A. Shulman-Peleg, M.Shatsky, R.Nussinov and H.J.Wolfson. Prediction of Interacting Single-Stranded RNA Bases by Protein-Binding Patterns. Journal of Molecular Biology, 379(2):299-316, 2008.

  • A. Shulman-Peleg, M.Shatsky, R.Nussinov and H.J.Wolfson. Spatial chemical conservation of hot spot interactions in protein-protein complexes, BMC Biology, 5(1):43, 2007

  • D. Zhi, M. Shatsky, S. Brenner. Alignment-Free Local Structural Search by Writhe Decomposition. (abstract) Workshop on Algorithms in Bioinformatics, Springer Verlag, Lecture Notes in Computer Science, 2007.

  • K. Lasker, O. Dror, M.Shatsky, R.Nussinov, H.Wolfson. EMatch: Discovery of High Resolution Structural Homologues of Protein Domains In Intermediate Resolution Cryo-EM. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 4(1):28-39, 2007.

  • M. Shatsky, A. Shulman-Peleg, R. Nussinov, H. Wolfson. The Multiple Common Point Set Problem and its Application to Molecule Binding Pattern Detection. Journal of Computational Biology, 13(2):407-28, 2006.

  • M. Shatsky, R. Nussinov, and H.J. Wolfson. Optimization of Multiple Sequence Alignment Based on Multiple Structure Alignment. Proteins: Structure, Function, and Bioinformatics, 62(1):209-17, 2006.

  • A. Shulman-Peleg, M. Shatsky, R. Nussinov, H. Wolfson. MAPPIS: Multiple 3D Alignment of Protein-Protein Interfaces. M.R. Berthold et al. (Eds.): CompLife 2005, Lecture Notes in Computer Science, Volume 3695, pp. 91-103, 2005.

  • M. Shatsky, A. Shulman-Peleg, R. Nussinov, H. Wolfson. Recognition of Binding Patterns Common to a Set of Protein Structures. RECOMB 2005, Lecture Notes in Computer Science, Springer, vol. 3500, pp. 440-455, 2005.

  • M. Shatsky, R. Nussinov, H. Wolfson. A Method for Simultaneous Alignment of Multiple Protein Structures. Proteins: Structure, Function, and Genetics, 56(1), 143-156, 2004.

  • M. Shatsky, R. Nussinov, H. Wolfson. FlexProt: Alignment of Flexible Protein Structures Without a Pre-definition of Hinge Regions. Journal of Computational Biology, 11(1), 83-106, 2004.

  • D. Schneidman-Duhovny, Y. Inbar, V. Polak, M. Shatsky, I. Halperin, H. Benyamini, A. Barzilay, O. Shem-Tov, N. Haspel, R. Nussinov, H. J. Wolfson. Taking Geometry to its Edge: Fast Unbound Rigid (and Hinge-bent) Docking. Proteins: Structure, Function, and Genetics, 52(1), 107-112, 2003.

  • M. Shatsky, H.J. Wolfson, R.Nussinov. MultiProt - a Multiple Protein Structural Alignment Algorithm. Workshop on Algorithms in Bioinformatics, Springer Verlag, Lecture Notes in Computer Science 2452: 235-250, 2002.

  • M. Shatsky, R. Nussinov, H. Wolfson. Flexible protein alignment and hinge detection. Proteins: Structure, Function, and Genetics,48:242-256, 2002.

  • B. Ma, M. Shatsky, H.J. Wolfson, R.Nussinov. Multiple diverse ligands binding at a single protein site: A matter of pre-existing populations. Protein Sci., 11, 184-197, 2002.

  • M. Shatsky, Z. Fligelman, R. Nussinov, H.J. Wolfson. Alignment of Flexible Protein Structures. Proc. 8th International Conference on Intelligent Systems for Molecular Biology (ISMB'00), 329-343, The AAAI press, 2000.

Book Chapters and Reviews:

  • M. Shatsky, R. Nussinov, and H.J. Wolfson. Algorithms for Multiple Protein Structure Alignment and Structure-Derived Multiple Sequence Alignment. Protein Structure Prediction Series: Methods in Molecular Biology , Vol. 413. Zaki, Mohammed; Bystroff, Chris (Eds.), 2nd ed., 2007. Springer, Humana Press.

  • H. Wolfson, M. Shatsky, D. Duhovny, O. Dror, A. Shulman, B.Ma, R. Nussinov. From structure to function: methods and applications. Curr Protein Pept Sci., 6(2):171-83, 2005.

Software/Web-servers:

  • A. Shulman-Peleg, M.Shatsky, R.Nussinov and H.J.Wolfson. MultiBind and MAPPIS: webservers for multiple alignment of protein 3D-binding sites and their interactions. Nucleic Acids Res. 36(Web Server issue):W260-4, 2008.

  • M. Shatsky, O. Dror, D. Duhovny, A. Shulman, R. Nussinov, H. Wolfson. BioInfo3D: A Suite of Tools for Structural Bioinformatics. Nucleic Acid Research 32, 503-507, 2004.

PhD thesis: download here.