Biomedical Engineering Reference
In-Depth Information
Figure 25.6
BLAST interface of GIPSY.
25.3.1.2 FASTA
Fast Alignment Search Tool (FASTA) is an algorithm capa-
ble of providing the best local optimal alignments of genomic sequences using the
combined heuristic and dynamic programming methods. The FASTA algorithm ini-
tially finds the regions shared by a pair of sequences having high density of identities,
subsequently rescanning the top 10 best scoring regions using scoring matrices like
PAM. A joining penalty can be given to connect the high-scoring regions, whereas
an optimal alignment is obtained in the final step by using a dynamic programming
method. Though the FASTA algorithm is comparatively slower than BLAST, the sen-
sitivity in finding high-ranking scores among distantly related sequences is higher.
FASTA, which has immense applications in studying protein functions, becomes
more compute-intensive especially for longer query sequences (more than 1 MB)
when searched against large databases like human genome sequence database, non-
redundant (nr) database and the like. Parallel FASTA programs were obtained from
the Virginia University and were ported and optimized on PARAM Padma. The
FASTA interface in GIPSY has programs, namely FASTA (protein against protein
or nucleotide against nucleotide database search), FASTX (nucleotide against pro-
tein), TFASTX (translated nucleotide against translated nucleotide database), and
TFASTY (protein against nucleotide) (Fig. 25.7). The sequence databases to search
and the scoring matrices to choose are the same as that of the BLAST interface. The
options that are available for users are K-tuple value, histogram (whether to display
or not), Expect value, and maximum score. Default values are used if the users do not
select any options. Similar to BLAST PSE, client-side validations are done in FASTA
PSE also.
Search WWH ::
Custom Search