Shotgun Sequencing - Whole Genome Shotgun Sequencing

Whole Genome Shotgun Sequencing

Whole genome shotgun sequencing for small (4000 to 7000 basepair) genomes was already in use in 1979. Broader application benefited from pairwise end sequencing, known colloquially as double-barrel shotgun sequencing. As sequencing projects began to take on longer and more complicated DNAs, multiple groups began to realize that useful information could be obtained by sequencing both ends of a fragment of DNA. Although sequencing both ends of the same fragment and keeping track of the paired data was more cumbersome than sequencing a single end of two distinct fragments, the knowledge that the two sequences were oriented in opposite directions and were about the length of a fragment apart from each other was valuable in reconstructing the sequence of the original target fragment. The first published description of the use of paired ends was in 1990 as part of the sequencing of the human HGPRT locus, although the use of paired ends was limited to closing gaps after the application of a traditional shotgun sequencing approach. The first theoretical description of a pure pairwise end sequencing strategy, assuming fragments of constant length, was in 1991. At the time, there was community consensus that the optimal fragment length for pairwise end sequencing would be three times the sequence read length. In 1995 Roach et al. introduced the innovation of using fragments of varying sizes, and demonstrated that a pure pairwise end-sequencing strategy would be possible on large targets. The strategy was subsequently adopted by The Institute for Genomic Research (TIGR) to sequence the genome of the bacterium Haemophilus influenzae in 1995, and then by Celera Genomics to sequence the Drosophila melanogaster (fruit fly) genome in 2000, and subsequently the human genome.

To apply the strategy, high-molecular-weight DNA is sheared into random fragments, size-selected (usually 2, 10, 50, and 150 kb), and cloned into an appropriate vector. The clones are then sequenced from both ends using the chain termination method yielding two short sequences. Each sequence is called an end-read or read and two reads from the same clone are referred to as mate pairs. Since the chain termination method usually can only produce reads between 500 and 1000 bases long, in all but the smallest clones, mate pairs will rarely overlap.

The original sequence is reconstructed from the reads using sequence assembly software. First, overlapping reads are collected into longer composite sequences known as contigs. Contigs can be linked together into scaffolds by following connections between mate pairs. The distance between contigs can be inferred from the mate pair positions if the average fragment length of the library is known and has a narrow window of deviation. Depending on the size of the gap between contigs, different techniques can be used to find the sequence in the gaps. If the gap is small (5-20kb) then the use of PCR to amplify the region is required, followed by sequencing. If the gap is large (>20kb) then the large fragment is cloned in special vectors such as BAC (Bacterial artificial chromosomes) followed by sequencing of the vector.

Proponents of this approach argue that it is possible to sequence the whole genome at once using large arrays of sequencers, which makes the whole process much more efficient than more traditional approaches. Detractors argue that although the technique quickly sequences large regions of DNA, its ability to correctly link these regions is suspect, particularly for genomes with repeating regions. As sequence assembly programs become more sophisticated and computing power becomes cheaper, it may be possible to overcome this limitation.