Assembly of DNA Sequencing Data

Abstract

At this writing, public databases contain the completed, contiguous sequences of a large number of bacterial genomes (e.g., http://www.tigr.org/tdb/mdb/mdbcomplete.html) (1), the yeast Saccharomyces cerevisiae genome (2); and the genomes of the nematode (3), Drosophila (4), and Arabidopsis (5). Public sequencing projects for many other genomes, including several large ones, are in progress. In the case of the publicly funded human genome sequencing effort, more than 45% of the >3-Gb genome is in finished form, including several completed chromosomes, with the remaining 55% expected to be finished by the spring of 2003. The term finished to describe sequence has special meaning and significance in the public genome-sequencing arena, where there is a general agreement that a large-insert clone, chromosome, or genome is considered finished when the accuracy of the sequence exceeds 99.99% (i.e., <1 error in 10,000 bp) and all gaps that can be filled by known techniques are filled, so that the sequence is either completely contiguous or has very few, well-annotated gaps.