Scientific data visualization and digital image processing for structural biology
Abstract
This thesis focuses on the design and development of algorithms and tools for interactive data visualization and digital image processing of large data sets produced in structural biology experiments. We describe various computational and visualization algorithms, which we have developed and implemented as part of the Tonitza package for interactive visualization and analysis of structural biology data. The computational algorithms include methods for fitting of data sets using correlation and scaling, various types of interpolation, and algorithms for generating statistical information. Several two- and three-dimensional representations of the data sets are described in detail. We present a number of image processing methods for extracting information from images, and we discuss their applications to electron microscopy. Such methods extend the scientist's ability to study images of biological structures. We describe the Crosspoint Method, a new technique we developed for automatic detection of the positions of virus particles in electron micrographs. We present the heuristics and the algorithms involved and compare the results obtained with those reported in the literature. We introduce EMMA, a new package for digital processing of micrograph images which includes the Crosspoint Method. As part of the research presented in this thesis, we also describe several parallel algorithms and load balancing schemes for structural biology applications. They helped improve our understanding of the complexity of the problems involved and of how the data is transformed from the moment it is collected until the moment it is displayed.
Degree
Ph.D.
Advisors
Marinescu, Purdue University.
Subject Area
Computer science|Biology|Cellular biology|Microbiology
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.