<s>
In	O
bioinformatics	O
,	O
a	O
sequence	B-Algorithm
alignment	I-Algorithm
is	O
a	O
way	O
of	O
arranging	O
the	O
sequences	O
of	O
DNA	O
,	O
RNA	O
,	O
or	O
protein	O
to	O
identify	O
regions	O
of	O
similarity	O
that	O
may	O
be	O
a	O
consequence	O
of	O
functional	O
,	O
structural	O
,	O
or	O
evolutionary	O
relationships	O
between	O
the	O
sequences	O
.	O
</s>
<s>
Aligned	O
sequences	O
of	O
nucleotide	O
or	O
amino	O
acid	O
residues	O
are	O
typically	O
represented	O
as	O
rows	O
within	O
a	O
matrix	B-Architecture
.	O
</s>
<s>
Sequence	B-Algorithm
alignments	I-Algorithm
are	O
also	O
used	O
for	O
non-biological	O
sequences	O
,	O
such	O
as	O
calculating	O
the	O
distance	O
cost	O
between	O
strings	O
in	O
a	O
natural	O
language	O
or	O
in	O
financial	O
data	O
.	O
</s>
<s>
In	O
sequence	B-Algorithm
alignments	I-Algorithm
of	O
proteins	O
,	O
the	O
degree	O
of	O
similarity	O
between	O
amino	O
acids	O
occupying	O
a	O
particular	O
position	O
in	O
the	O
sequence	O
can	O
be	O
interpreted	O
as	O
a	O
rough	O
measure	O
of	O
how	O
conserved	O
a	O
particular	O
region	O
or	O
sequence	O
motif	O
is	O
among	O
lineages	O
.	O
</s>
<s>
Instead	O
,	O
human	O
knowledge	O
is	O
applied	O
in	O
constructing	O
algorithms	O
to	O
produce	O
high-quality	O
sequence	B-Algorithm
alignments	I-Algorithm
,	O
and	O
occasionally	O
in	O
adjusting	O
the	O
final	O
results	O
to	O
reflect	O
patterns	O
that	O
are	O
difficult	O
to	O
represent	O
algorithmically	O
(	O
especially	O
in	O
the	O
case	O
of	O
nucleotide	O
sequences	O
)	O
.	O
</s>
<s>
Computational	O
approaches	O
to	O
sequence	B-Algorithm
alignment	I-Algorithm
generally	O
fall	O
into	O
two	O
categories	O
:	O
global	O
alignments	O
and	O
local	B-Algorithm
alignments	I-Algorithm
.	O
</s>
<s>
By	O
contrast	O
,	O
local	B-Algorithm
alignments	I-Algorithm
identify	O
regions	O
of	O
similarity	O
within	O
long	O
sequences	O
that	O
are	O
often	O
widely	O
divergent	O
overall	O
.	O
</s>
<s>
Local	B-Algorithm
alignments	I-Algorithm
are	O
often	O
preferable	O
,	O
but	O
can	O
be	O
more	O
difficult	O
to	O
calculate	O
because	O
of	O
the	O
additional	O
challenge	O
of	O
identifying	O
the	O
regions	O
of	O
similarity	O
.	O
</s>
<s>
A	O
variety	O
of	O
computational	O
algorithms	O
have	O
been	O
applied	O
to	O
the	O
sequence	B-Algorithm
alignment	I-Algorithm
problem	O
.	O
</s>
<s>
These	O
include	O
slow	O
but	O
formally	O
correct	O
methods	O
like	O
dynamic	B-Algorithm
programming	I-Algorithm
.	O
</s>
<s>
These	O
also	O
include	O
efficient	O
,	O
heuristic	B-Algorithm
algorithms	I-Algorithm
or	O
probabilistic	O
methods	O
designed	O
for	O
large-scale	O
database	O
search	O
,	O
that	O
do	O
not	O
guarantee	O
to	O
find	O
best	O
matches	O
.	O
</s>
<s>
In	O
almost	O
all	O
sequence	B-Algorithm
alignment	I-Algorithm
representations	O
,	O
sequences	O
are	O
written	O
in	O
rows	O
arranged	O
so	O
that	O
aligned	O
residues	O
appear	O
in	O
successive	O
columns	O
.	O
</s>
<s>
For	O
multiple	O
sequences	O
the	O
last	O
row	O
in	O
each	O
column	O
is	O
often	O
the	O
consensus	O
sequence	O
determined	O
by	O
the	O
alignment	O
;	O
the	O
consensus	O
sequence	O
is	O
also	O
often	O
represented	O
in	O
graphical	O
format	O
with	O
a	O
sequence	B-Application
logo	I-Application
in	O
which	O
the	O
size	O
of	O
each	O
nucleotide	O
or	O
amino	O
acid	O
letter	O
corresponds	O
to	O
its	O
degree	O
of	O
conservation	O
.	O
</s>
<s>
Sequence	B-Algorithm
alignments	I-Algorithm
can	O
be	O
stored	O
in	O
a	O
wide	O
variety	O
of	O
text-based	O
file	O
formats	O
,	O
many	O
of	O
which	O
were	O
originally	O
developed	O
in	O
conjunction	O
with	O
a	O
specific	O
alignment	B-Application
program	I-Application
or	O
implementation	O
.	O
</s>
<s>
Most	O
web-based	O
tools	O
allow	O
a	O
limited	O
number	O
of	O
input	O
and	O
output	O
formats	O
,	O
such	O
as	O
FASTA	B-Application
format	O
and	O
GenBank	O
format	O
and	O
the	O
output	O
is	O
not	O
easily	O
editable	O
.	O
</s>
<s>
Several	O
conversion	O
programs	O
that	O
provide	O
graphical	O
and/or	O
command	O
line	O
interfaces	O
are	O
available	O
,	O
such	O
as	O
and	O
EMBOSS	B-Application
.	O
</s>
<s>
There	O
are	O
also	O
several	O
programming	O
packages	O
which	O
provide	O
this	O
conversion	O
functionality	O
,	O
such	O
as	O
BioPython	B-Library
,	O
BioRuby	B-Application
and	O
BioPerl	B-Application
.	O
</s>
<s>
A	O
general	O
global	O
alignment	O
technique	O
is	O
the	O
Needleman	B-Algorithm
–	I-Algorithm
Wunsch	I-Algorithm
algorithm	I-Algorithm
,	O
which	O
is	O
based	O
on	O
dynamic	B-Algorithm
programming	I-Algorithm
.	O
</s>
<s>
Local	B-Algorithm
alignments	I-Algorithm
are	O
more	O
useful	O
for	O
dissimilar	O
sequences	O
that	O
are	O
suspected	O
to	O
contain	O
regions	O
of	O
similarity	O
or	O
similar	O
sequence	O
motifs	O
within	O
their	O
larger	O
sequence	O
context	O
.	O
</s>
<s>
The	O
Smith	B-Algorithm
–	I-Algorithm
Waterman	I-Algorithm
algorithm	I-Algorithm
is	O
a	O
general	O
local	B-Algorithm
alignment	I-Algorithm
method	O
based	O
on	O
the	O
same	O
dynamic	B-Algorithm
programming	I-Algorithm
scheme	O
but	O
with	O
additional	O
choices	O
to	O
start	O
and	O
end	O
at	O
any	O
place	O
.	O
</s>
<s>
In	O
this	O
case	O
,	O
neither	O
global	O
nor	O
local	B-Algorithm
alignment	I-Algorithm
is	O
entirely	O
appropriate	O
:	O
a	O
global	O
alignment	O
would	O
attempt	O
to	O
force	O
the	O
alignment	O
to	O
extend	O
beyond	O
the	O
region	O
of	O
overlap	O
,	O
while	O
a	O
local	B-Algorithm
alignment	I-Algorithm
might	O
not	O
fully	O
cover	O
the	O
region	O
of	O
overlap	O
.	O
</s>
<s>
Fast	O
expansion	O
of	O
genetic	O
data	O
challenges	O
speed	O
of	O
current	O
DNA	O
sequence	B-Algorithm
alignment	I-Algorithm
algorithms	O
.	O
</s>
<s>
Pairwise	O
sequence	B-Algorithm
alignment	I-Algorithm
methods	O
are	O
used	O
to	O
find	O
the	O
best-matching	O
piecewise	O
(	O
local	O
or	O
global	O
)	O
alignments	O
of	O
two	O
query	O
sequences	O
.	O
</s>
<s>
The	O
three	O
primary	O
methods	O
of	O
producing	O
pairwise	O
alignments	O
are	O
dot-matrix	O
methods	O
,	O
dynamic	B-Algorithm
programming	I-Algorithm
,	O
and	O
word	O
methods	O
;	O
however	O
,	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
techniques	O
can	O
also	O
align	O
pairs	O
of	O
sequences	O
.	O
</s>
<s>
One	O
way	O
of	O
quantifying	O
the	O
utility	O
of	O
a	O
given	O
pairwise	O
alignment	O
is	O
the	O
'	O
maximal	B-Algorithm
unique	I-Algorithm
match	I-Algorithm
 '	O
(	O
MUM	O
)	O
,	O
or	O
the	O
longest	O
subsequence	O
that	O
occurs	O
in	O
both	O
query	O
sequences	O
.	O
</s>
<s>
in	O
the	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
of	O
genomes	O
in	O
computational	O
biology	O
.	O
</s>
<s>
Identification	O
of	O
MUMs	O
and	O
other	O
potential	O
anchors	O
,	O
is	O
the	O
first	O
step	O
in	O
larger	O
alignment	O
systems	O
such	O
as	O
MUMmer	B-Application
.	O
</s>
<s>
thumb|200px|A	O
DNA	O
dot	B-Application
plot	I-Application
of	O
a	O
human	O
zinc	O
finger	O
transcription	O
factor	O
(	O
GenBank	O
ID	O
NM_002383	O
)	O
,	O
showing	O
regional	O
self-similarity	O
.	O
</s>
<s>
The	O
main	B-Algorithm
diagonal	I-Algorithm
represents	O
the	O
sequence	O
's	O
alignment	O
with	O
itself	O
;	O
lines	O
off	O
the	O
main	B-Algorithm
diagonal	I-Algorithm
represent	O
similar	O
or	O
repetitive	O
patterns	O
within	O
the	O
sequence	O
.	O
</s>
<s>
This	O
is	O
a	O
typical	O
example	O
of	O
a	O
recurrence	B-Application
plot	I-Application
.	O
</s>
<s>
The	O
dot-matrix	O
approach	O
,	O
which	O
implicitly	O
produces	O
a	O
family	O
of	O
alignments	O
for	O
individual	O
sequence	O
regions	O
,	O
is	O
qualitative	O
and	O
conceptually	O
simple	O
,	O
though	O
time-consuming	O
to	O
analyze	O
on	O
a	O
large	O
scale	O
.	O
</s>
<s>
In	O
the	O
absence	O
of	O
noise	O
,	O
it	O
can	O
be	O
easy	O
to	O
visually	O
identify	O
certain	O
sequence	O
features	O
—	O
such	O
as	O
insertions	O
,	O
deletions	O
,	O
repeats	O
,	O
or	O
inverted	O
repeats	O
—	O
from	O
a	O
dot-matrix	B-Application
plot	I-Application
.	O
</s>
<s>
To	O
construct	O
a	O
dot-matrix	B-Application
plot	I-Application
,	O
the	O
two	O
sequences	O
are	O
written	O
along	O
the	O
top	O
row	O
and	O
leftmost	O
column	O
of	O
a	O
two-dimensional	O
matrix	B-Architecture
and	O
a	O
dot	O
is	O
placed	O
at	O
any	O
point	O
where	O
the	O
characters	O
in	O
the	O
appropriate	O
columns	O
match	O
—	O
this	O
is	O
a	O
typical	O
recurrence	B-Application
plot	I-Application
.	O
</s>
<s>
The	O
dot	B-Application
plots	I-Application
of	O
very	O
closely	O
related	O
sequences	O
will	O
appear	O
as	O
a	O
single	O
line	O
along	O
the	O
matrix	B-Architecture
's	O
main	B-Algorithm
diagonal	I-Algorithm
.	O
</s>
<s>
Problems	O
with	O
dot	B-Application
plots	I-Application
as	O
an	O
information	O
display	O
technique	O
include	O
:	O
noise	O
,	O
lack	O
of	O
clarity	O
,	O
non-intuitiveness	O
,	O
difficulty	O
extracting	O
match	O
summary	O
statistics	O
and	O
match	O
positions	O
on	O
the	O
two	O
sequences	O
.	O
</s>
<s>
Dot	B-Application
plots	I-Application
can	O
also	O
be	O
used	O
to	O
assess	O
repetitiveness	O
in	O
a	O
single	O
sequence	O
.	O
</s>
<s>
A	O
sequence	O
can	O
be	O
plotted	O
against	O
itself	O
and	O
regions	O
that	O
share	O
significant	O
similarities	O
will	O
appear	O
as	O
lines	O
off	O
the	O
main	B-Algorithm
diagonal	I-Algorithm
.	O
</s>
<s>
The	O
technique	O
of	O
dynamic	B-Algorithm
programming	I-Algorithm
can	O
be	O
applied	O
to	O
produce	O
global	O
alignments	O
via	O
the	O
Needleman-Wunsch	B-Algorithm
algorithm	I-Algorithm
,	O
and	O
local	B-Algorithm
alignments	I-Algorithm
via	O
the	O
Smith-Waterman	B-Algorithm
algorithm	I-Algorithm
.	O
</s>
<s>
In	O
typical	O
usage	O
,	O
protein	O
alignments	O
use	O
a	O
substitution	B-Algorithm
matrix	I-Algorithm
to	O
assign	O
scores	O
to	O
amino-acid	O
matches	O
or	O
mismatches	O
,	O
and	O
a	O
gap	O
penalty	O
for	O
matching	O
an	O
amino	O
acid	O
in	O
one	O
sequence	O
to	O
a	O
gap	O
in	O
the	O
other	O
.	O
</s>
<s>
DNA	O
and	O
RNA	O
alignments	O
may	O
use	O
a	O
scoring	O
matrix	B-Architecture
,	O
but	O
in	O
practice	O
often	O
simply	O
assign	O
a	O
positive	O
match	O
score	O
,	O
a	O
negative	O
mismatch	O
score	O
,	O
and	O
a	O
negative	O
gap	O
penalty	O
.	O
</s>
<s>
(	O
In	O
standard	O
dynamic	B-Algorithm
programming	I-Algorithm
,	O
the	O
score	O
of	O
each	O
amino	O
acid	O
position	O
is	O
independent	O
of	O
the	O
identity	O
of	O
its	O
neighbors	O
,	O
and	O
therefore	O
base	O
stacking	O
effects	O
are	O
not	O
taken	O
into	O
account	O
.	O
</s>
<s>
Dynamic	B-Algorithm
programming	I-Algorithm
can	O
be	O
useful	O
in	O
aligning	O
nucleotide	O
to	O
protein	O
sequences	O
,	O
a	O
task	O
complicated	O
by	O
the	O
need	O
to	O
take	O
into	O
account	O
frameshift	O
mutations	O
(	O
usually	O
insertions	O
or	O
deletions	O
)	O
.	O
</s>
<s>
Its	O
ability	O
to	O
evaluate	O
frameshifts	O
offset	O
by	O
an	O
arbitrary	O
number	O
of	O
nucleotides	O
makes	O
the	O
method	O
useful	O
for	O
sequences	O
containing	O
large	O
numbers	O
of	O
indels	O
,	O
which	O
can	O
be	O
very	O
difficult	O
to	O
align	O
with	O
more	O
efficient	O
heuristic	B-Algorithm
methods	O
.	O
</s>
<s>
In	O
practice	O
,	O
the	O
method	O
requires	O
large	O
amounts	O
of	O
computing	O
power	O
or	O
a	O
system	O
whose	O
architecture	O
is	O
specialized	O
for	O
dynamic	B-Algorithm
programming	I-Algorithm
.	O
</s>
<s>
The	O
BLAST	B-Application
and	O
EMBOSS	B-Application
suites	O
provide	O
basic	O
tools	O
for	O
creating	O
translated	O
alignments	O
(	O
though	O
some	O
of	O
these	O
approaches	O
take	O
advantage	O
of	O
side-effects	O
of	O
sequence	O
searching	O
capabilities	O
of	O
the	O
tools	O
)	O
.	O
</s>
<s>
More	O
general	O
methods	O
are	O
available	O
from	O
open-source	B-Application
software	I-Application
such	O
as	O
.	O
</s>
<s>
The	O
dynamic	B-Algorithm
programming	I-Algorithm
method	O
is	O
guaranteed	O
to	O
find	O
an	O
optimal	O
alignment	O
given	O
a	O
particular	O
scoring	O
function	O
;	O
however	O
,	O
identifying	O
a	O
good	O
scoring	O
function	O
is	O
often	O
an	O
empirical	O
rather	O
than	O
a	O
theoretical	O
matter	O
.	O
</s>
<s>
Although	O
dynamic	B-Algorithm
programming	I-Algorithm
is	O
extensible	O
to	O
more	O
than	O
two	O
sequences	O
,	O
it	O
is	O
prohibitively	O
slow	O
for	O
large	O
numbers	O
of	O
sequences	O
or	O
extremely	O
long	O
sequences	O
.	O
</s>
<s>
Word	O
methods	O
,	O
also	O
known	O
as	O
k-tuple	O
methods	O
,	O
are	O
heuristic	B-Algorithm
methods	O
that	O
are	O
not	O
guaranteed	O
to	O
find	O
an	O
optimal	O
alignment	O
solution	O
,	O
but	O
are	O
significantly	O
more	O
efficient	O
than	O
dynamic	B-Algorithm
programming	I-Algorithm
.	O
</s>
<s>
Word	O
methods	O
are	O
best	O
known	O
for	O
their	O
implementation	O
in	O
the	O
database	O
search	O
tools	O
FASTA	B-Application
and	O
the	O
BLAST	B-Application
family	O
.	O
</s>
<s>
In	O
the	O
FASTA	B-Application
method	O
,	O
the	O
user	O
defines	O
a	O
value	O
k	O
to	O
use	O
as	O
the	O
word	O
length	O
with	O
which	O
to	O
search	O
the	O
database	O
.	O
</s>
<s>
The	O
BLAST	B-Application
family	O
of	O
search	O
methods	O
provides	O
a	O
number	O
of	O
algorithms	O
optimized	O
for	O
particular	O
types	O
of	O
queries	O
,	O
such	O
as	O
searching	O
for	O
distantly	O
related	O
sequence	O
matches	O
.	O
</s>
<s>
BLAST	B-Application
was	O
developed	O
to	O
provide	O
a	O
faster	O
alternative	O
to	O
FASTA	B-Application
without	O
sacrificing	O
much	O
accuracy	O
;	O
like	O
FASTA	B-Application
,	O
BLAST	B-Application
uses	O
a	O
word	O
search	O
of	O
length	O
k	O
,	O
but	O
evaluates	O
only	O
the	O
most	O
significant	O
word	O
matches	O
,	O
rather	O
than	O
every	O
word	O
match	O
as	O
does	O
FASTA	B-Application
.	O
</s>
<s>
Most	O
BLAST	B-Application
implementations	O
use	O
a	O
fixed	O
default	O
word	O
length	O
that	O
is	O
optimized	O
for	O
the	O
query	O
and	O
database	O
type	O
,	O
and	O
that	O
is	O
changed	O
only	O
under	O
special	O
circumstances	O
,	O
such	O
as	O
when	O
searching	O
with	O
repetitive	O
or	O
very	O
short	O
query	O
sequences	O
.	O
</s>
<s>
Multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
is	O
an	O
extension	O
of	O
pairwise	O
alignment	O
to	O
incorporate	O
more	O
than	O
two	O
sequences	O
at	O
a	O
time	O
.	O
</s>
<s>
Multiple	O
sequence	B-Algorithm
alignments	I-Algorithm
are	O
computationally	O
difficult	O
to	O
produce	O
and	O
most	O
formulations	O
of	O
the	O
problem	O
lead	O
to	O
NP-complete	O
combinatorial	O
optimization	O
problems	O
.	O
</s>
<s>
The	O
technique	O
of	O
dynamic	B-Algorithm
programming	I-Algorithm
is	O
theoretically	O
applicable	O
to	O
any	O
number	O
of	O
sequences	O
;	O
however	O
,	O
because	O
it	O
is	O
computationally	O
expensive	O
in	O
both	O
time	O
and	O
memory	B-General_Concept
,	O
it	O
is	O
rarely	O
used	O
for	O
more	O
than	O
three	O
or	O
four	O
sequences	O
in	O
its	O
most	O
basic	O
form	O
.	O
</s>
<s>
This	O
method	O
requires	O
constructing	O
the	O
n-dimensional	O
equivalent	O
of	O
the	O
sequence	O
matrix	B-Architecture
formed	O
from	O
two	O
sequences	O
,	O
where	O
n	O
is	O
the	O
number	O
of	O
sequences	O
in	O
the	O
query	O
.	O
</s>
<s>
Standard	O
dynamic	B-Algorithm
programming	I-Algorithm
is	O
first	O
used	O
on	O
all	O
pairs	O
of	O
query	O
sequences	O
and	O
then	O
the	O
"	O
alignment	O
space	O
"	O
is	O
filled	O
in	O
by	O
considering	O
possible	O
matches	O
or	O
gaps	O
at	O
intermediate	O
positions	O
,	O
eventually	O
constructing	O
an	O
alignment	O
essentially	O
between	O
each	O
two-sequence	O
alignment	O
.	O
</s>
<s>
One	O
method	O
for	O
reducing	O
the	O
computational	O
demands	O
of	O
dynamic	B-Algorithm
programming	I-Algorithm
,	O
which	O
relies	O
on	O
the	O
"	O
sum	O
of	O
pairs	O
"	O
objective	O
function	O
,	O
has	O
been	O
implemented	O
in	O
the	O
software	O
package	O
.	O
</s>
<s>
Progressive	O
,	O
hierarchical	O
,	O
or	O
tree	O
methods	O
generate	O
a	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
by	O
first	O
aligning	O
the	O
most	O
similar	O
sequences	O
and	O
then	O
adding	O
successively	O
less	O
related	O
sequences	O
or	O
groups	O
to	O
the	O
alignment	O
until	O
the	O
entire	O
query	O
set	O
has	O
been	O
incorporated	O
into	O
the	O
solution	O
.	O
</s>
<s>
The	O
initial	O
tree	O
describing	O
the	O
sequence	O
relatedness	O
is	O
based	O
on	O
pairwise	O
comparisons	O
that	O
may	O
include	O
heuristic	B-Algorithm
pairwise	O
alignment	O
methods	O
similar	O
to	O
FASTA	B-Application
.	O
</s>
<s>
Most	O
progressive	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
methods	O
additionally	O
weight	O
the	O
sequences	O
in	O
the	O
query	O
set	O
according	O
to	O
their	O
relatedness	O
,	O
which	O
reduces	O
the	O
likelihood	O
of	O
making	O
a	O
poor	O
choice	O
of	O
initial	O
sequences	O
and	O
thus	O
improves	O
alignment	O
accuracy	O
.	O
</s>
<s>
Many	O
variations	O
of	O
the	O
Clustal	B-Application
progressive	O
implementation	O
are	O
used	O
for	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
,	O
phylogenetic	O
tree	O
construction	O
,	O
and	O
as	O
input	O
for	O
protein	O
structure	O
prediction	O
.	O
</s>
<s>
A	O
slower	O
but	O
more	O
accurate	O
variant	O
of	O
the	O
progressive	O
method	O
is	O
known	O
as	O
T-Coffee	B-Application
.	O
</s>
<s>
The	O
realigned	O
subsets	O
are	O
then	O
themselves	O
aligned	O
to	O
produce	O
the	O
next	O
iteration	O
's	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
.	O
</s>
<s>
Motif	O
finding	O
,	O
also	O
known	O
as	O
profile	O
analysis	O
,	O
constructs	O
global	O
multiple	O
sequence	B-Algorithm
alignments	I-Algorithm
that	O
attempt	O
to	O
align	O
short	O
conserved	O
sequence	O
motifs	O
among	O
the	O
sequences	O
in	O
the	O
query	O
set	O
.	O
</s>
<s>
This	O
is	O
usually	O
done	O
by	O
first	O
constructing	O
a	O
general	O
global	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
,	O
after	O
which	O
the	O
highly	O
conserved	O
regions	O
are	O
isolated	O
and	O
used	O
to	O
construct	O
a	O
set	O
of	O
profile	O
matrices	O
.	O
</s>
<s>
The	O
profile	O
matrix	B-Architecture
for	O
each	O
conserved	O
region	O
is	O
arranged	O
like	O
a	O
scoring	O
matrix	B-Architecture
but	O
its	O
frequency	O
counts	O
for	O
each	O
amino	O
acid	O
or	O
nucleotide	O
at	O
each	O
position	O
are	O
derived	O
from	O
the	O
conserved	O
region	O
's	O
character	O
distribution	O
rather	O
than	O
from	O
a	O
more	O
general	O
empirical	O
distribution	O
.	O
</s>
<s>
In	O
cases	O
where	O
the	O
original	O
data	B-General_Concept
set	I-General_Concept
contained	O
a	O
small	O
number	O
of	O
sequences	O
,	O
or	O
only	O
highly	O
related	O
sequences	O
,	O
pseudocounts	O
are	O
added	O
to	O
normalize	O
the	O
character	O
distributions	O
represented	O
in	O
the	O
motif	O
.	O
</s>
<s>
A	O
variety	O
of	O
general	O
optimization	O
algorithms	O
commonly	O
used	O
in	O
computer	O
science	O
have	O
also	O
been	O
applied	O
to	O
the	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
problem	O
.	O
</s>
<s>
Hidden	O
Markov	O
models	O
have	O
been	O
used	O
to	O
produce	O
probability	O
scores	O
for	O
a	O
family	O
of	O
possible	O
multiple	O
sequence	B-Algorithm
alignments	I-Algorithm
for	O
a	O
given	O
query	O
set	O
;	O
although	O
early	O
HMM-based	O
methods	O
produced	O
underwhelming	O
performance	O
,	O
later	O
applications	O
have	O
found	O
them	O
especially	O
effective	O
in	O
detecting	O
remotely	O
related	O
sequences	O
because	O
they	O
are	O
less	O
susceptible	O
to	O
noise	O
created	O
by	O
conservative	O
or	O
semiconservative	O
substitutions	O
.	O
</s>
<s>
Genetic	B-Algorithm
algorithms	I-Algorithm
and	O
simulated	B-Algorithm
annealing	I-Algorithm
have	O
also	O
been	O
used	O
in	O
optimizing	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
scores	O
as	O
judged	O
by	O
a	O
scoring	O
function	O
like	O
the	O
sum-of-pairs	O
method	O
.	O
</s>
<s>
More	O
complete	O
details	O
and	O
software	O
packages	O
can	O
be	O
found	O
in	O
the	O
main	O
article	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
.	O
</s>
<s>
The	O
Burrows	B-Algorithm
–	I-Algorithm
Wheeler	I-Algorithm
transform	I-Algorithm
has	O
been	O
successfully	O
applied	O
to	O
fast	O
short	O
read	O
alignment	O
in	O
popular	O
tools	O
such	O
as	O
Bowtie	B-Application
and	O
BWA	O
.	O
</s>
<s>
See	O
FM-index	B-Algorithm
.	O
</s>
<s>
These	O
methods	O
can	O
be	O
used	O
for	O
two	O
or	O
more	O
sequences	O
and	O
typically	O
produce	O
local	B-Algorithm
alignments	I-Algorithm
;	O
however	O
,	O
because	O
they	O
depend	O
on	O
the	O
availability	O
of	O
structural	O
information	O
,	O
they	O
can	O
only	O
be	O
used	O
for	O
sequences	O
whose	O
corresponding	O
structures	O
are	O
known	O
(	O
usually	O
through	O
X-ray	O
crystallography	O
or	O
NMR	O
spectroscopy	O
)	O
.	O
</s>
<s>
The	O
DALI	O
method	O
,	O
or	O
distance	O
matrix	B-Architecture
alignment	O
,	O
is	O
a	O
fragment-based	O
method	O
for	O
constructing	O
structural	O
alignments	O
based	O
on	O
contact	O
similarity	O
patterns	O
between	O
successive	O
hexapeptides	O
in	O
the	O
query	O
sequences	O
.	O
</s>
<s>
SSAP	O
(	O
sequential	O
structure	O
alignment	B-Application
program	I-Application
)	O
is	O
a	O
dynamic	O
programming-based	O
method	O
of	O
structural	O
alignment	O
that	O
uses	O
atom-to-atom	O
vectors	O
in	O
structure	O
space	O
as	O
comparison	O
points	O
.	O
</s>
<s>
Based	O
on	O
measures	O
such	O
as	O
rigid-body	O
root	O
mean	O
square	O
distance	O
,	O
residue	O
distances	O
,	O
local	O
secondary	O
structure	O
,	O
and	O
surrounding	O
environmental	O
features	O
such	O
as	O
residue	O
neighbor	O
hydrophobicity	O
,	O
local	B-Algorithm
alignments	I-Algorithm
called	O
"	O
aligned	O
fragment	O
pairs	O
"	O
are	O
generated	O
and	O
used	O
to	O
build	O
a	O
similarity	O
matrix	B-Architecture
representing	O
all	O
possible	O
structural	O
alignments	O
within	O
predefined	O
cutoff	O
criteria	O
.	O
</s>
<s>
A	O
path	O
from	O
one	O
protein	O
structure	O
state	O
to	O
the	O
other	O
is	O
then	O
traced	O
through	O
the	O
matrix	B-Architecture
by	O
extending	O
the	O
growing	O
alignment	O
one	O
fragment	O
at	O
a	O
time	O
.	O
</s>
<s>
Phylogenetics	O
and	O
sequence	B-Algorithm
alignment	I-Algorithm
are	O
closely	O
related	O
fields	O
due	O
to	O
the	O
shared	O
necessity	O
of	O
evaluating	O
sequence	O
relatedness	O
.	O
</s>
<s>
The	O
field	O
of	O
phylogenetics	O
makes	O
extensive	O
use	O
of	O
sequence	B-Algorithm
alignments	I-Algorithm
in	O
the	O
construction	O
and	O
interpretation	O
of	O
phylogenetic	O
trees	O
,	O
which	O
are	O
used	O
to	O
classify	O
the	O
evolutionary	O
relationships	O
between	O
homologous	O
genes	O
represented	O
in	O
the	O
genomes	O
of	O
divergent	O
species	O
.	O
</s>
<s>
Roughly	O
speaking	O
,	O
high	O
sequence	B-Algorithm
identity	I-Algorithm
suggests	O
that	O
the	O
sequences	O
in	O
question	O
have	O
a	O
comparatively	O
young	O
most	O
recent	O
common	O
ancestor	O
,	O
while	O
low	O
identity	O
suggests	O
that	O
the	O
divergence	O
is	O
more	O
ancient	O
.	O
</s>
<s>
This	O
approximation	O
,	O
which	O
reflects	O
the	O
"	O
molecular	O
clock	O
"	O
hypothesis	O
that	O
a	O
roughly	O
constant	O
rate	O
of	O
evolutionary	O
change	O
can	O
be	O
used	O
to	O
extrapolate	O
the	O
elapsed	O
time	O
since	O
two	O
genes	O
first	O
diverged	O
(	O
that	O
is	O
,	O
the	O
coalescence	O
time	O
)	O
,	O
assumes	O
that	O
the	O
effects	O
of	O
mutation	O
and	O
selection	B-Application
are	O
constant	O
across	O
sequence	O
lineages	O
.	O
</s>
<s>
Other	O
techniques	O
that	O
assemble	O
multiple	O
sequence	B-Algorithm
alignments	I-Algorithm
and	O
phylogenetic	O
trees	O
score	O
and	O
sort	O
trees	O
first	O
and	O
calculate	O
a	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
from	O
the	O
highest-scoring	O
tree	O
.	O
</s>
<s>
Commonly	O
used	O
methods	O
of	O
phylogenetic	O
tree	O
construction	O
are	O
mainly	O
heuristic	B-Algorithm
because	O
the	O
problem	O
of	O
selecting	O
the	O
optimal	O
tree	O
,	O
like	O
the	O
problem	O
of	O
selecting	O
the	O
optimal	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
,	O
is	O
NP-hard	O
.	O
</s>
<s>
Sequence	B-Algorithm
alignments	I-Algorithm
are	O
useful	O
in	O
bioinformatics	O
for	O
identifying	O
sequence	O
similarity	O
,	O
producing	O
phylogenetic	O
trees	O
,	O
and	O
developing	O
homology	O
models	O
of	O
protein	O
structures	O
.	O
</s>
<s>
However	O
,	O
the	O
biological	O
relevance	O
of	O
sequence	B-Algorithm
alignments	I-Algorithm
is	O
not	O
always	O
clear	O
.	O
</s>
<s>
In	O
database	O
searches	O
such	O
as	O
BLAST	B-Application
,	O
statistical	O
methods	O
can	O
determine	O
the	O
likelihood	O
of	O
a	O
particular	O
alignment	O
between	O
sequences	O
or	O
sequence	O
regions	O
arising	O
by	O
chance	O
given	O
the	O
size	O
and	O
composition	O
of	O
the	O
database	O
being	O
searched	O
.	O
</s>
<s>
Repetitive	O
sequences	O
in	O
the	O
database	O
or	O
query	O
can	O
also	O
distort	O
both	O
the	O
search	O
results	O
and	O
the	O
assessment	O
of	O
statistical	O
significance	O
;	O
BLAST	B-Application
automatically	O
filters	O
such	O
repetitive	O
sequences	O
in	O
the	O
query	O
to	O
avoid	O
apparent	O
hits	O
that	O
are	O
statistical	O
artifacts	O
.	O
</s>
<s>
Methods	O
of	O
statistical	O
significance	O
estimation	O
for	O
gapped	O
sequence	B-Algorithm
alignments	I-Algorithm
are	O
available	O
in	O
the	O
literature	O
.	O
</s>
<s>
Methods	O
of	O
alignment	O
credibility	O
estimation	O
for	O
gapped	O
sequence	B-Algorithm
alignments	I-Algorithm
are	O
available	O
in	O
the	O
literature	O
.	O
</s>
<s>
Protein	O
sequences	O
are	O
frequently	O
aligned	O
using	O
substitution	B-Algorithm
matrices	I-Algorithm
that	O
reflect	O
the	O
probabilities	O
of	O
given	O
character-to-character	O
substitutions	O
.	O
</s>
<s>
Another	O
common	O
series	O
of	O
scoring	O
matrices	O
,	O
known	O
as	O
BLOSUM	O
(	O
Blocks	O
Substitution	B-Algorithm
Matrix	I-Algorithm
)	O
,	O
encodes	O
empirically	O
derived	O
substitution	O
probabilities	O
.	O
</s>
<s>
Variants	O
of	O
both	O
types	O
of	O
matrices	O
are	O
used	O
to	O
detect	O
sequences	O
with	O
differing	O
levels	O
of	O
divergence	O
,	O
thus	O
allowing	O
users	O
of	O
BLAST	B-Application
or	O
FASTA	B-Application
to	O
restrict	O
searches	O
to	O
more	O
closely	O
related	O
matches	O
or	O
expand	O
to	O
detect	O
more	O
divergent	O
sequences	O
.	O
</s>
<s>
It	O
can	O
be	O
very	O
useful	O
and	O
instructive	O
to	O
try	O
the	O
same	O
alignment	O
several	O
times	O
with	O
different	O
choices	O
for	O
scoring	O
matrix	B-Architecture
and/or	O
gap	O
penalty	O
values	O
and	O
compare	O
the	O
results	O
.	O
</s>
<s>
Sequence	B-Algorithm
alignment	I-Algorithm
is	O
also	O
a	O
part	O
of	O
genome	O
assembly	O
,	O
where	O
sequences	O
are	O
aligned	O
to	O
find	O
overlap	O
so	O
that	O
contigs	B-Application
(	O
long	O
stretches	O
of	O
sequence	O
)	O
can	O
be	O
formed	O
.	O
</s>
<s>
The	O
methods	O
used	O
for	O
biological	O
sequence	B-Algorithm
alignment	I-Algorithm
have	O
also	O
found	O
applications	O
in	O
other	O
fields	O
,	O
most	O
notably	O
in	O
natural	B-Language
language	I-Language
processing	I-Language
and	O
in	O
social	O
sciences	O
,	O
where	O
the	O
Needleman-Wunsch	B-Algorithm
algorithm	I-Algorithm
is	O
usually	O
referred	O
to	O
as	O
Optimal	O
matching	O
.	O
</s>
<s>
Techniques	O
that	O
generate	O
the	O
set	O
of	O
elements	O
from	O
which	O
words	O
will	O
be	O
selected	O
in	O
natural-language	O
generation	O
algorithms	O
have	O
borrowed	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
techniques	O
from	O
bioinformatics	O
to	O
produce	O
linguistic	O
versions	O
of	O
computer-generated	O
mathematical	O
proofs	O
.	O
</s>
<s>
In	O
the	O
field	O
of	O
historical	O
and	O
comparative	O
linguistics	O
,	O
sequence	B-Algorithm
alignment	I-Algorithm
has	O
been	O
used	O
to	O
partially	O
automate	O
the	O
comparative	O
method	O
by	O
which	O
linguists	O
traditionally	O
reconstruct	O
languages	O
.	O
</s>
<s>
Business	O
and	O
marketing	O
research	O
has	O
also	O
applied	O
multiple	O
sequence	B-Algorithm
alignment	I-Algorithm
techniques	O
in	O
analyzing	O
series	O
of	O
purchases	O
over	O
time	O
.	O
</s>
<s>
A	O
more	O
complete	O
list	O
of	O
available	O
software	O
categorized	O
by	O
algorithm	O
and	O
alignment	O
type	O
is	O
available	O
at	O
sequence	B-Application
alignment	I-Application
software	I-Application
,	O
but	O
common	O
software	O
tools	O
used	O
for	O
general	O
sequence	B-Algorithm
alignment	I-Algorithm
tasks	O
include	O
ClustalW2	O
and	O
T-coffee	B-Application
for	O
alignment	O
,	O
and	O
BLAST	B-Application
and	O
FASTA3x	O
for	O
database	O
searching	O
.	O
</s>
<s>
Commercial	O
tools	O
such	O
as	O
DNASTAR	O
Lasergene	O
,	O
Geneious	O
,	O
and	O
PatternHunter	B-Application
are	O
also	O
available	O
.	O
</s>
<s>
Alignment	O
algorithms	O
and	O
software	O
can	O
be	O
directly	O
compared	O
to	O
one	O
another	O
using	O
a	O
standardized	O
set	O
of	O
benchmark	O
reference	O
multiple	O
sequence	B-Algorithm
alignments	I-Algorithm
known	O
as	O
BAliBASE	O
.	O
</s>
<s>
The	O
data	B-General_Concept
set	I-General_Concept
consists	O
of	O
structural	O
alignments	O
,	O
which	O
can	O
be	O
considered	O
a	O
standard	O
against	O
which	O
purely	O
sequence-based	O
methods	O
are	O
compared	O
.	O
</s>
