... representations¹

also called fragments (see Myers (1995)) or readings (reads)

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... accurately²

e.g. for new electrophoresis methods

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...mira³

which is an acronym for Mimicking Intelligent Read Assembly

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... work.⁴

The automatic editor is subject of a thesis to be presented by Thomas Pfisterer

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... character.⁵

e.g., the symbol 'W' for an uncertainty between A (Adenosin) and T (Thymin)

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... orientation⁶

the experimentally gained sequences have a 50% chance of being in reverse complement orientation

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... traces⁷

Mean length of useful sequences gathered on ABI 3730 machines at The Institute of Genomic Research (TIGR) in 2003, pers. communication from Bill Niermann (Investigator at TIGR) in April 2004

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... bases.⁸

insertions and deletions are commonly referred to as indels

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... reads''.⁹

The term contig is derived from ``contiguous sequence''

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... Green¹⁰

PHRAP as acronym for PHils Revised Assembly Program, see also http://www.phrap.org/

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...fuguization¹¹

this refers to the rather compact genome of the puffer fish (Fugu rubripes) which is largely devoid of large copy repeats.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... enough.¹²

e.g. for new electrophoresis methods

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... operations¹³

Specialised hardware for this type of operations starts approximately at EUR 500,000

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... information¹⁴

which they call ``double-barreled data''

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... mind'¹⁵

e.g. 'could the base G at position 235 in read 4 be replaced by a A?' (because the overall consensus at this position of the other reads suggests this possibility)

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... task.¹⁶

For example, quality clipping, sequencing vector and cosmid vector removal can be controlled by the PREGAP4 environment provided with the GAP4 package (Bonfield et al. (1995b); Staden (1996); Bonfield and Staden (1996)) or the LUCY program, parts of these tasks can also be done with cross_match provided by the PHRAP package or other packages like, e.g., PFP from Paracel (Paracel (2002a)).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... biosciences¹⁷

see also Myers (1991)

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... sequence¹⁸

It is again of no consequence which sequence is in reverse complement direction to the other, as both will be searched with the reverse complement pattern of the other one.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... ZEBRA¹⁹

ZEBRA is not an acronym, but the algorithm was named because it produces 'bands' in memory which resemble the patterns of the african Zebra

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... number²⁰

which is 4^k for k Ns.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... sequences.²¹

see also the paper from Pearson (1998) for a review on empirical statistical estimates for sequence similarity searches

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... exemplarily.²²

The graphical display of the resulting bit vectors reminds vaguely to the african zebra, hence the algorithm's name.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... devised²³

for example see Grice et al. (1997); Chao et al. (1994) for an overview and Chao et al. (1995) for an application

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... mismatches.²⁴

but perhaps one or several aligns of a base against an 'N'

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... reads²⁵

Chimeric reads, as described in section 2.2.2, must be considered as garbage.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... contigs.²⁶

Of course, a single read itself cannot be called a contig. But putting it into the same data structure (a contig object) like the other, assembled reads is a convenient way to keep unassembled reads in a database.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... projects²⁷

cosmid, BAC or even whole genome size

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... project.²⁸

for example to close gaps

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... occurring.²⁹

e.g. the infamous AG-problem known with the ABI 373 and 377 machines where a G preceeded by an A is often unincisive

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... positions.³⁰

although chemistry together with the sequencing direction of a read might play a minor role on the type of errors generated, but this has no real impact on the error distribution itself

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... threshold³¹

working with relatively high score ratios beginning with 80%

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... routines.³²

This is one of the more prominent places where it shows that the EST assembler is a sibling of its genome pendant.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...mira³³

MIRA: Mimicking Intelligent Read Assembly

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... sequences³⁴

see Notredame (2002) for a review of state-of-the-art algorithms,Thompson et al. (1999a) and Lassmann and Sonnhammer (2002) for a evaluation of some of these tools and Morgenstern et al. (2003) for the description of a web-based solution

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... GAP4/cycle³⁵

GAP4/cycle is a script performing several GAP4 assemblies with decreasing strictness

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... test.³⁶

PHRAP uses base qualities for the assembly, MIRA/EdIt can use them if present and GAP4/cycle does not

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... 2.0.1³⁷

TraceTuner is from from Paracel Inc.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... capabilities³⁸

It is widely known that the speed of a program depends mainly on the quality of the algorithms it bases on. However, good compilers can squeeze a considerable amount of execution speed from optimal algorithms by optimising them on machine level, sometimes to a factor of 3 and more.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... EGCS³⁹

which later on was promoted as official GCC until the new GCC 3 compiler lineage appeared by mid of 2001

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... .o''⁴⁰

mind the blank between the asterisk and .o

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... hierarchy⁴¹

especially into a single-rooted hierarchy

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.