DNA sequence analysis and the computer.

Abstract

Analysis of DNA sequences has greatly aided our understanding of molecular biology and genetics. Nucleotide sequences have provided a wealth of information about gene structure and function. In addition to revealing the primary structure of genes, sequencing has uncovered the phenomena of overlapping genes, segmented genes, pseudogenes and a different genetic code for mitochondria. To date over 2 million base pairs of DNA sequence have been determined, nearly all during the last 9 years. In this short article I shall briefly discuss the rapid shotgun methods of sequence analysis and the computer facilities required for data collation and analysis. The original sequencing strategies relied upon the possession of a detailed restriction map of the target DNA, in order that a planned campaign could be devised. However, the construction of such a restriction map takes time, in most cases as long as it would take to sequence the actual DNA. It quickly became obvious that it was easier to sequence fragments indiscriminately, from several restriction-enzyme digests, as the sequence of the DNA yields the ultimate in restriction maps. The most laborious aspect of these earlier methods was the isolation and purification of suitable restriction enzyme fragments, from agarose or polyacrylamide gels. A major innovation was achieved by the use of a biological cloning procedure, to replace the electrophoretic separation of restriction fragments. This results in a very efficient purification procedure and is affected by neither the complexity of the restriction enzyme digest, nor the size of the resulting fragments. Suitable cloning vectors have been developed, which allow the direct sequencing of the cloned DNA by either the Sanger polymerase copying methods(Sanger & Coulson, 1975; Sangeretal., 1977)or the chemical degradation procedure of Maxam & Gilbert (1977). Chemical sequencing vectors utilize a polylinker of unique restriction enzyme sites, which allow the cloning and subsequent sequence analysis of the DNA, without recourse to purification of end-labelled fragments. Analysis can thus be performed on plasmid DNA isolated from small cultures,by quick minipreparative methods (Ruther, 1982). The Maxam & Gilbert (1977) procedure involves many chemical manipulations, which make the process both tedious and timeconsuming. The drudgery of these steps may soon be vanquished in laboratories which can afford to purchase the microchemical sequencing robot, developed recently in Japan (Wada, 1984). The initial disadvantage of the Sanger chain-termination method (Sanger et al., 1977), was the requirement of singlestranded DNA as a template for the polymerase reaction. This problem was elegantly overcome by the introduction of filamentous bacteriophage cloning vectors (Messing et al., 1977; Zinder & Boeke, 1982). These single-stranded DNA phages, owing to their mode of replication and their life cycle, offer a quick and easy means of DNA purification. Sufficient DNA for analysis can be obtained from as little as 1 ml cultures of infected cells. These ease and simplicity of the Sanger et al. (1977) method, coupled with the efficacy of the Messing vectors (Messing et al., 1981), makes the M13 shotgun sequencing system the method of choice for DNA sequence analysis. Recent advances in the methodology employed have allowed for the rapid accumulation of sequence data. The complete DNA sequences of bacteriophage lambda, 48 502 base-pairs (Sanger et al., 1982), and the human herpes virus, Epstein Barr Virus, approximately 175 OOO base-pairs (B. G. Barrell, personal communication), have recently been determined using this system. Improved gel technology