Editing Shotgun sequencing (section)

==Hierarchical shotgun sequencing==
[[File:Whole genome shotgun sequencing versus Hierarchical shotgun sequencing.png|thumb|In whole genome shotgun sequencing (top), the entire genome is sheared randomly into small fragments (appropriately sized for sequencing) and then reassembled. In hierarchical shotgun sequencing (bottom), the genome is first broken into larger segments. After the order of these segments is deduced, they are further sheared into fragments appropriately sized for sequencing.]]

Although shotgun sequencing can in theory be applied to a genome of any size, its direct application to the sequencing of large genomes (for instance, the [[human genome]]) was limited until the late 1990s, when technological advances made practical the handling of the vast quantities of complex data involved in the process.<ref name="genome sequencing">{{cite journal |last1=Dunham |first1=Ian |title=Genome Sequencing |journal=Encyclopedia of Life Sciences |date=9 September 2005 |doi=10.1038/npg.els.0005378|isbn=978-0-470-01617-6 }}</ref> Historically, full-genome shotgun sequencing was believed to be limited by both the sheer size of large genomes and by the complexity added by the high percentage of repetitive DNA (greater than 50% for the human genome) present in large genomes.<ref name="venter">{{cite journal |last1=Venter |first1=J Craig |title=Shotgunning the Human Genome: A Personal View |journal=Encyclopedia of Life Sciences |date=9 September 2005 |doi=10.1038/npg.els.0005850|isbn=978-0-470-01617-6 }}</ref> It was not widely accepted that a full-genome shotgun sequence of a large genome would provide reliable data. For these reasons, other strategies that lowered the computational load of sequence assembly had to be utilized before shotgun sequencing was performed.<ref name="venter" />
In hierarchical sequencing, also known as top-down sequencing, a low-resolution [[gene mapping#Physical mapping|physical map]] of the genome is made prior to actual sequencing. From this map, a minimal number of fragments that cover the entire chromosome are selected for sequencing.<ref name="textbook">Gibson, G. and Muse, S. V. ''A Primer of Genome Science''. 3rd ed. P.84</ref> In this way, the minimum amount of high-throughput sequencing and assembly is required.

The amplified genome is first sheared into larger pieces (50-200kb) and cloned into a bacterial host using BACs or [[P1-derived artificial chromosome]]s (PAC). Because multiple genome copies have been sheared at random, the fragments contained in these clones have different ends, and with enough coverage (see section above) finding the smallest possible '''scaffold''' of [[contig#BAC contigs|BAC contigs]] that covers the entire genome is theoretically possible. This scaffold is called the '''minimum tiling path'''.[[File:Tiling path.png|thumb|A BAC contig that covers the entire genomic area of interest makes up the tiling path.]] Once a tiling path has been found, the BACs that form this path are sheared at random into smaller fragments and can be sequenced using the shotgun method on a smaller scale.<ref>{{Cite journal |last1=Bozdag |first1=Serdar |last2=Close |first2=Timothy J. |last3=Lonardi |first3=Stefano |date=March 2013 |title=A Graph-Theoretical Approach to the Selection of the Minimum Tiling Path from a Physical Map |journal=IEEE/ACM Transactions on Computational Biology and Bioinformatics |volume=10 |issue=2 |pages=352–360 |doi=10.1109/tcbb.2013.26 |pmid=23929859 |issn=1545-5963}}</ref>

Although the full sequences of the BAC contigs is not known, their orientations relative to one another are known. There are several methods for deducing this order and selecting the BACs that make up a tiling path. The general strategy involves identifying the positions of the clones relative to one another and then selecting the fewest clones required to form a contiguous scaffold that covers the entire area of interest. The order of the clones is deduced by determining the way in which they overlap.<ref name="genome map">{{cite journal |last1=Dear |first1=Paul H |title=Genome Mapping |journal=Encyclopedia of Life Sciences |date=9 September 2005 |doi=10.1038/npg.els.0005353|isbn=978-0-470-01617-6 }}</ref> Overlapping clones can be identified in several ways. A small radioactively or chemically labeled probe containing a [[sequence-tagged site]] (STS) can be hybridized onto a microarray upon which the clones are printed.<ref name="genome map" /> In this way, all the clones that contain a particular sequence in the genome are identified. The end of one of these clones can then be sequenced to yield a new probe and the process repeated in a method called chromosome walking.

Alternatively, the [[BAC library#Genomic libraries|BAC library]] can be [[restriction digest|restriction-digested]]. Two clones that have several fragment sizes in common are inferred to overlap because they contain multiple similarly spaced restriction sites in common.<ref name="genome map" /> This method of genomic mapping is called restriction or BAC fingerprinting because it identifies a set of restriction sites contained in each clone. Once the overlap between the clones has been found and their order relative to the genome known, a scaffold of a minimal subset of these contigs that covers the entire genome is shotgun-sequenced.<ref name="textbook" />

Because it involves first creating a low-resolution map of the genome, hierarchical shotgun sequencing is slower than whole-genome shotgun sequencing, but relies less heavily on computer algorithms than whole-genome shotgun sequencing. The process of extensive BAC library creation and tiling path selection, however, make hierarchical shotgun sequencing slow and labor-intensive. Now that the technology is available and the reliability of the data demonstrated,<ref name="venter" /> the speed and cost efficiency of whole-genome shotgun sequencing has made it the primary method for genome sequencing.