ABSTRACT
Shotgun sequencing has been employed since the invention of rapid sequencing in 1976 and, although there are many methods that modify this essentially random approach, it is still widely used. The computer science aspects of sequencing by hybridization are quite distinct from that for shotgun sequencing. It is certain that othernew techniques will be developed or significantly refined to make genome sequencing routine. It is less certain what role mathematics or computer science will play, but the excitement of this important enterprise is contagious. It is natural to ask how accurate assembled DNA sequences are. Most of the information about this important topic comes from sequence that is determined in different laboratories and then compared. Still, sequence assembly is basically a random process and we should be able to make a statistical statement about the finished product. A reasonable stochastic model for sequence assembly is closelyrelated to the ocean and islands model for physical mapping by fingerprinting random clones.
