BLAST, the Basic
Local Alignment Search Tool, is a statistical
pattern matching algorithm. It was developed
and published by Altshul
et al. in 1990 and then as an enhanced
version in 1997. It is one of the foundational
algorithms for the study of comparative genomics.
BLAST's impact on our understanding of biology
is demonstrated by its ubiquity. BLAST is
web-base and fast. It is used world-wide to
compare DNA and protein sequences for similarity
in structure and function and to infer evolutionary
relationships between sequences. As an example
of the volume of BLAST analyses conducted
worldwide, in March 2003, the US National
Center for Biotechnology Information (NCBI)
was receiving 100,000 unique BLAST runs from
70,000 unique IP addresses daily, with usage
increasing continually. ( Personal communication
W. Matten, 2003.)
BLAST operates by cutting up sequences
into smaller "words" and searching
for each of the words in "target"
sequences. It looks in both directions
along target sequences to find longer
pattern matches. BLAST scores matches
according to experimental knowledge of
homology. This accounts for some of the
imperfect matches it generates. BLAST
also matches and aligns sequences locally.
It does not create global sequence alignments.
BLAST ranks resulting matches according
to the likelihood they are homologues.To
learn more about BLAST visit the NCBI
Tutorial.
Ecce
Homology visualizes the BLAST algorithm
as it operates on protein sequences.
To the right is an image of a human gene
character from the installation undergoing
BLAST. The human gene (translated into
protein) selected from the characters
on the vertical axis is enlarged in the
central area where the viewer’s
gesture traces had been. The collection
of points at the upper left represents
the query sequence being segmented into
“words” that are compared
to the target database sequences depicted
on the horizontal axis. BLAST results
are depicted as human gene characters
overlaid upon target organism gene characters.
The ubiquitous
use of BLAST arose in part to its lightning
fast speed and ease of accessibility over
web-based servers. This has progressed to
the point where today one can submit sequences
in bulk to web-based BLAST servers and go
attend to something else while the sequences
are analyzed. For Ecce
Homology we have slowed the process
down to a scale of time that human beings
can observe. We show the operation of BLAST
by visualizing some of the algorithms intermediary
processes. In this way, BLAST drives the aesthetic
experience for visitors in real-time while
it is comparing a human gene (translated to
its protein product) to one from the rice
plant. As part of this we have designed the
installation to incorporate an aesthetic
of slowness.
