Bettina Keszthelyi

post doctorate

Department of Biochemistry & Biophysics
University of California, San Francisco
600 16th St
San Francisco CA
94143-2240
(415) 502-4316 (Ph)
(415) 476-1902 (Fax)

New Computational Methods for EM Tomography

My recent interest is in the 3D study of cellular structures using EM tomography. I received my PhD in theoretical particle physics in 1991 and continued doing theoretical particle and particle astrophysics research as a postdoctoral fellow at the University of Wisconsin, Madison and at the University of North Carolina, Chapel Hill before I joined the Agard Lab.

Computerized EM tomography makes possible the 3D study of cellular structures, like centrosomes and chromosomes. The basic idea is that the 3D structure of an object can be reconstructed from many 2D views (projections) taken by the electron microscope using some backprojection method. Although the simplest such method, called weighted backprojection, is efficient and fast, there is plenty of room for improvement. I have developed a software package which includes an iterative post reconstruction algorithm that addresses most of the questions related to inconsistent, incomplete data. Further improvements will include:
(1) The deconvolution of the contrast transfer function (CTF) of the microscope, the experimental mapping and the deconvolution of the point spread function of the CCD camera.
(2) Taking into account the multiple scattering of the electrons within the specimen, especially for higher tilts and thick specimens, and possible divergences of the beam.
(3) Developing a more accurate algorithm for the alignment of the projection data, which is based on cross-correlation techniques, and does not requires gold bead markers.

Because the above described iterative alignment and reconstruction can take days on a single processor, the development of the parallel version of the software is underway in collaboration with the National Center for Supercomputing Applications (NCSA) at the University of Illinois, Urbana.

I am also interested in developing new computation methods, which are faster and make it possible to recover more information from the data. Fractional Fourier transforms and Periodic Discrete Radon Transformations are two posibilities I am already considering.

Publications:

Zheng SQ, Keszthelyi B, Branlund E, Lyle JM, Braunfeld MB, Sedat JW, and Agard DA,"UCSF tomography: An integrated software suite for real-time electron microscopic tomographic data collection, alignment, and reconstruction," J Struct Biol. 2006 Jun 23 (html or pdf).