Biology Reference
In-Depth Information
Table 2
BLAST-specific programs
BLAST program Query
Library Comments
Protein Protein Highly sensitive iterative protein similarity searches [
8
]
psiblast
Protein PSSMs Protein searches against a PSSM library of conserved domains [
8
]
rpsblast
DNA
DNA More sensitive DNA:DNA comparison at the translated
DNA level [
6
,
8
]
tblastx
DNA
DNA
(
blastn -task megablast
) high-speed DNA
matching for Near-identical sequences (e.g. ESTs vs genomes).
The default for
blastn
megablast
translated DNA searches return 10-50-times as many homologs as
DNA:DNA searches, with much more reliable statistical estimates.
Both the BLAST
1
and FASTA software packages offer different
programs for different similarity searching programs, depending
on the source of the query sequences. Table
1
summarizes the
programs in the BLAST and FASTA packages that have very similar
functions; Table
2
lists programs found only in the BLAST package;
Table
3
lists FASTA-specific programs.
Both packages offer heuristic strategies for rapid sequence
comparison (
blast
,
fasta
). For more sensitive searches the
BLAST package provides
psiblast
and
rpsblast
, which use
models of protein families encoded as Position Specific Scoring
Matrices. Almost all the BLAST package programs use some ver-
sion of the heuristic approach to calculate
local
similarity scores
implemented by the BLAST algorithm. BLAST package typically
requires specially formatted databases, created with
makeblastdb
,
except when only two sequences are compared.
2
In addition to heuristic methods, the FASTA package offers
a variety of optimal search algorithms, including the local Smith
-Waterman algorithm [
2
] implemented in
ssearch
, as well as
optimal global alignment algorithms (
ggsearch
and
glsearch
).
An implementation of the
psisearch
strategy that uses the opti-
mal
ssearch
program is available at the European Bioinformatics
Institute (
www.ebi.ac.uk/Tools/sss
,[
3
]). FASTA also provides
the
lalign
program, which uses the
sim
[
4
] implementation
1
The programs listed in Tables
1
and
2
are part of the NCBI BLAST+ distribution [
1
]. An earlier version of the
BLAST distribution used the
blastall
program with the
-p blastp
option to specify the specific
program.
2
Previous versions of BLAST provided
bl2seq
to compare two sequences in FASTA format. The BLAST
programs now provide the “Blast2Sequences” mode by using
-subject
option, rather than the
-db
option.
