"Frontmatter". In: Plant Genomics and Proteomics

G
ETTING A
C
OMPLETE
G
ENOME
S
EQUENCE
Two alternative, but not mutually exclusive, sequencing strategies for the
whole genome have been developed (Figure 3.1). The first, as used for deci-
phering the Arabidopsis sequence and the IRGSP rice sequencing effort, is one
in which an anchored physical map is generated and then the sequencing is
done on all the parts that comprise the physical assembly. The second strat-
egy essentially eschews the need for a prior physical map and generates
enormous quantities of random sequence data that are then bioinformati-
cally assembled into a physical genome of some degree of completeness. This
second strategy is the shotgun strategy that was used by the Syngenta and
Beijing rice genome sequencing projects. It is clearly applicable for a simple
genome where every sequence is represented once, or at most a few times,
so that a limited number of assemblies can be completed. It has been suc-
cessfully applied to generate the complete sequence of many microbes.
However, the assembly task becomes much more difficult when there are
many copies of dispersed repeated sequences, which themselves can be
complex. Indeed, there is a difference of opinion as to whether or not the
private (Celera) shotgun human sequence could have been assembled into
large regions without the publicly available physical map (Butcher, 2001).
The problem with plant genomes becomes even more pronounced because
the evolutionary history indicates multiple rounds of polyploidy leaving
related regions in very different parts of the genome. The highly repetitive
nature of the maize genome, for example, which contains large numbers of
dispersed very similar repeats, may well be impossible to assemble correctly
into a complete genome sequence with the current bioinformatics tools. The
assembly of the two rice shotgun sequences each resulted in tens of thou-
sands of contiguous fragments of length less than 7 kb, and the rice genome
is only one-fifth the size of the maize genome (Goff et al., 2002). 
G
E T T I N G A
C
O M P L E T E
G
E N O M E
S
E Q U E N C E
4 9

---