BIOINFORMATICS<-->STRUCTURE
Jerusalem, Israel, November 17-21, 1996

Abstract


Approaches to highly automated genome sequencing

F. William Studier, John J. Dunn, Janine S. Graves, Jan Kieleczawa, Sean R. McCorkle, Mark A. Quesada, Keith H. Thompson and Shiping Zhang

Biology Department, Brookhaven National Laboratory, Upton, NY 11790, USA

studier@genome1.bio.bnl.gov


We are developing technology aimed at improving the accuracy, speed and efficiency of genome sequencing. As a test bed, we are sequencing the 1-Mbp linear chromosome of Borrelia burgdorferi, the bacterium that causes Lyme disease. The strategy is to use a limited initial shotgun phase followed by directed 2nd-end sequencing to create a framework that is filled in by primer walking. Primers for walking by cycle sequencing with four-color fluorescent terminators are produced by ligation of hexamers on hexamer templates, selecting from the complete library of 4096 hexamers. Initial sequencing has used a commercial instrument, but a prototype module for rapid parallel analysis of 12 samples by capillary electrophoresis is nearing completion. Base calling software that provides a quality level for each base is being evaluated and refined by comparing its results with commercial base calling and human editing on the entire data set for this project. A major challenge is to develop software that uses the quality information to process, assemble and edit sequences, and to generate consensus sequences with a meaningful quality score at every base, largely or entirely without human intervention. Success would allow automated selection of primers in the walking phase and, ultimately, automation of the entire sequencing process.


Back to the Invited Speakers Index.