The genomes of some pathogenic viruses are made up of RNA that can be involved directly in theproduction of proteins crucial to viral reproduction and infection. These fast-mutating RNA viruses showed ahigh incidence of cross-infections among different species by variant forms (Plant et al. 2005). It is feasible toperform computational analyses on the viral genome and obtain useful information which give clues to theorigin, natural reservoir, and evolution of the virus, contributing to the understanding of the immune responseto these viruses and the pathogenesis of viral diseases and facilitate the development of antiviral drugs. Wehave recently reported that regions in coronavirus RNA genomes with statistically significant clusters ofpalindromes are associated with the presence of stem-loops or pseudoknots (Chew et al 2004). These RNAstructures have been suggested to be responsible for frame-shifting mechanisms during gene expression,where two different proteins can be produced in the same region just by shifting the reading frame by onebase. We hypothesize that by exploiting the correlation between nonrandom clusters of close inversions withstem-loop and pseudoknot structures in RNA molecules, an efficient and utilitarian tool for predictingsecondary structures on RNA viral genomes can be developed using current heterogeneous Grid Computingtechnology. We propose a four-year project to evaluate this hypothesis focusing on coronaviruses andinfluenza viruses with specific aims to: (1) Establish a statistics-based algorithm to locate RNA segmentcontaining nonrandom clusters of close inversions. (2) Develop a strategy for cutting of the viral genomesequences into segments of length no greater than 200 bases. (3) Construct a software prototype for RNAsecondary structure prediction using Grid Computing technology. (4) Implement user facilities for thesoftware and predict genome structures in coronaviruses and Influenza viruses. For this project, ourobjective is to produce the prototype of an RNA secondary structure prediction system on a grid ofheterogeneous, distributed computers. The software will be publicly accessible through a web portal. Ourlong-term goal is to develop a set of open-source computational tools in a Grid Computing environment toaccurately predict genome structures and dynamics in RNA viruses and their interactions with cellular RNA.This will provide information to be used by virologists to design finely tuned experiments to study RNAviruses and their pathogenic interactions with their hosts, especially when time is limiting in combating new infectious viral diseases.
Showing the most recent 10 out of 139 publications
