344
V. Mittard-Runte et al.
can be assigned to the unknown gene. A variety of computational tools for detecting sequence similarity are available with the Basic Local Alignment Search Tool
(BLAST) (Altschul et al. 1990) being one of the best known and most widespread.
9.3.4.2 Protein Domains
Protein sequences can also be analysed in detail for the presence of conserved
domains. Domains can be seen as the functional building blocks of proteins.
Specialized databases such as that constructed by the InterPro Consortium have
been established for this task (see Section 9.3.3.3).
Sequence analyses and similarity searches against databases produce evidence
that support the functional annotation of a gene. If the detected sequence similarity
is low or the evidence is contradictory, an assignment of the function of a gene is difficult and requires human interaction. Otherwise, a propagation of false annotations
may be introduced in the genomic databases. Functional annotation is in general
thought to be of best quality when performed by a human expert.
9.3.4.3 Use of Gene Clusters in Functional Annotation
As described before, sequence similarity can be used to determine the function of
genes. The results derived from sequence analyses can be enriched by incorporating
the evidence of higher-level analyses of the genomic data. The order of genes on
the chromosome of an organism often yields additional information about functionally related genes, particularly for prokaryotes. In bacteria the sequence of genes
on the chromosome (gene order) is well preserved at close phylogenetic distances
(Tamames et al. 1997), but the order and composition of genes in two diverging
organisms changes over time. Events like duplication or loss of genes as well as horizontal gene transfer change the composition of genes in the genome. Translocation,
transposition, inversion, and chromosome fission and fusion affect the gene order.
Interestingly, sets of genes with strong conservation of composition and order
can also be detected in distantly related species. The term “gene cluster” was used
for the first time by Bauerle and Margolin (1966) who described the tryptophan
gene cluster in Salmonella typhimurium. Through the detection of further occurrences of the tryptophan cluster in other organisms, Tatsuov et al. (1996) showed
in the mid-90s that, although gene order is generally under no selective pressure in
prokaryotic genomes, certain genes tend to conserve their chromosomal neighbourhood. This circumstance can be explained by evolutionary advantages that derive
from the chromosomal neighbourhood of the respective genes for the organisms.
Functionally related genes occur in close proximity for example, if they are collectively regulated in an operon in prokaryotic genomes. If their gene products interact,
it is advantageous for the cell to produce them at a close distance (Dandekar et al.
1998).
Another reason for conserved gene order across phylogenetic distances is the
occurrence of horizontal gene transfer (Lawrence and Roth 1996). It is reasonable
to use the information present in the chromosomal context of the gene as presented
V. Mittard-Runte et al.
can be assigned to the unknown gene. A variety of computational tools for detecting sequence similarity are available with the Basic Local Alignment Search Tool
(BLAST) (Altschul et al. 1990) being one of the best known and most widespread.
9.3.4.2 Protein Domains
Protein sequences can also be analysed in detail for the presence of conserved
domains. Domains can be seen as the functional building blocks of proteins.
Specialized databases such as that constructed by the InterPro Consortium have
been established for this task (see Section 9.3.3.3).
Sequence analyses and similarity searches against databases produce evidence
that support the functional annotation of a gene. If the detected sequence similarity
is low or the evidence is contradictory, an assignment of the function of a gene is difficult and requires human interaction. Otherwise, a propagation of false annotations
may be introduced in the genomic databases. Functional annotation is in general
thought to be of best quality when performed by a human expert.
9.3.4.3 Use of Gene Clusters in Functional Annotation
As described before, sequence similarity can be used to determine the function of
genes. The results derived from sequence analyses can be enriched by incorporating
the evidence of higher-level analyses of the genomic data. The order of genes on
the chromosome of an organism often yields additional information about functionally related genes, particularly for prokaryotes. In bacteria the sequence of genes
on the chromosome (gene order) is well preserved at close phylogenetic distances
(Tamames et al. 1997), but the order and composition of genes in two diverging
organisms changes over time. Events like duplication or loss of genes as well as horizontal gene transfer change the composition of genes in the genome. Translocation,
transposition, inversion, and chromosome fission and fusion affect the gene order.
Interestingly, sets of genes with strong conservation of composition and order
can also be detected in distantly related species. The term “gene cluster” was used
for the first time by Bauerle and Margolin (1966) who described the tryptophan
gene cluster in Salmonella typhimurium. Through the detection of further occurrences of the tryptophan cluster in other organisms, Tatsuov et al. (1996) showed
in the mid-90s that, although gene order is generally under no selective pressure in
prokaryotic genomes, certain genes tend to conserve their chromosomal neighbourhood. This circumstance can be explained by evolutionary advantages that derive
from the chromosomal neighbourhood of the respective genes for the organisms.
Functionally related genes occur in close proximity for example, if they are collectively regulated in an operon in prokaryotic genomes. If their gene products interact,
it is advantageous for the cell to produce them at a close distance (Dandekar et al.
1998).
Another reason for conserved gene order across phylogenetic distances is the
occurrence of horizontal gene transfer (Lawrence and Roth 1996). It is reasonable
to use the information present in the chromosomal context of the gene as presented