<s>
The	O
HH-suite	B-Application
is	O
an	O
open-source	B-Application
software	I-Application
package	O
for	O
sensitive	O
protein	O
sequence	O
searching	O
.	O
</s>
<s>
HHpred	O
is	O
an	O
online	O
server	O
for	O
protein	O
structure	O
prediction	O
that	O
uses	O
homology	O
information	O
from	O
HH-suite	B-Application
.	O
</s>
<s>
The	O
HH-suite	B-Application
searches	O
for	O
sequences	O
using	O
hidden	O
Markov	O
models	O
(	O
HMMs	O
)	O
.	O
</s>
<s>
Among	O
the	O
most	O
popular	O
methods	O
for	O
protein	O
sequence	O
matching	O
,	O
the	O
programs	O
have	O
been	O
cited	O
more	O
than	O
5000	O
times	O
total	O
according	O
to	O
Google	B-Library
Scholar	I-Library
.	O
</s>
<s>
The	O
HHpred	O
server	O
and	O
the	O
HH-suite	B-Application
software	O
package	O
offer	O
many	O
popular	O
,	O
regularly	O
updated	O
databases	O
,	O
such	O
as	O
the	O
Protein	O
Data	O
Bank	O
,	O
as	O
well	O
as	O
the	O
InterPro	O
,	O
Pfam	O
,	O
COG	O
,	O
and	O
SCOP	O
databases	O
.	O
</s>
<s>
They	O
may	O
be	O
used	O
to	O
compare	O
a	O
sequence	O
to	O
a	O
profile	O
,	O
or	O
in	O
more	O
advanced	O
cases	O
such	O
as	O
HH-suite	B-Application
,	O
to	O
match	O
among	O
profiles	O
.	O
</s>
<s>
Profiles	O
and	O
alignments	O
are	O
themselves	O
derived	O
from	O
matches	O
,	O
using	O
for	O
example	O
PSI-BLAST	B-Application
or	O
HHblits	O
.	O
</s>
<s>
the	O
position-specific	O
degree	O
of	O
conservation	O
)	O
,	O
profile-profile	O
comparison	O
methods	O
are	O
much	O
more	O
powerful	O
than	O
sequence-sequence	O
comparison	O
methods	O
like	O
BLAST	B-Application
or	O
profile-sequence	O
comparison	O
methods	O
like	O
PSI-BLAST	B-Application
.	O
</s>
<s>
The	O
databases	O
contain	O
HMMs	O
that	O
are	O
precalculated	O
in	O
the	O
same	O
fashion	O
using	O
PSI-BLAST	B-Application
.	O
</s>
<s>
HHblits	O
,	O
a	O
part	O
of	O
the	O
HH-suite	B-Application
since	O
2001	O
,	O
builds	O
high-quality	O
multiple	O
sequence	O
alignments	O
(	O
MSAs	O
)	O
starting	O
from	O
a	O
single	O
query	O
sequence	O
or	O
a	O
MSA	O
.	O
</s>
<s>
As	O
in	O
PSI-BLAST	B-Application
,	O
it	O
works	O
iteratively	O
,	O
repeatedly	O
constructing	O
new	O
query	O
profiles	O
by	O
adding	O
the	O
results	O
found	O
in	O
the	O
previous	O
round	O
.	O
</s>
<s>
The	O
HH-suite	B-Application
comes	O
with	O
a	O
number	O
of	O
pre-built	O
profile	O
HMMs	O
that	O
can	O
be	O
searched	O
using	O
HHblits	O
and	O
HHsearch	O
,	O
among	O
them	O
a	O
clustered	O
version	O
of	O
the	O
UniProt	O
database	O
,	O
of	O
the	O
Protein	O
Data	O
Bank	O
of	O
proteins	O
with	O
known	O
structures	O
,	O
of	O
Pfam	O
protein	O
family	O
alignments	O
,	O
of	O
SCOP	O
structural	O
protein	O
domains	O
,	O
and	O
many	O
more	O
.	O
</s>
<s>
If	O
a	O
significant	O
match	O
with	O
a	O
protein	O
of	O
known	O
structure	O
(	O
a	O
"	O
template	O
"	O
)	O
is	O
found	O
in	O
the	O
PDB	O
database	O
,	O
HHpred	O
allows	O
the	O
user	O
to	O
build	O
a	O
homology	O
model	O
using	O
the	O
MODELLER	B-Application
software	O
,	O
starting	O
from	O
the	O
pairwise	O
query-template	O
alignment	O
.	O
</s>
<s>
In	O
addition	O
to	O
HHsearch	O
and	O
HHblits	O
,	O
the	O
HH-suite	B-Application
contains	O
programs	O
and	O
perl	O
scripts	O
for	O
format	O
conversion	O
,	O
filtering	O
of	O
MSAs	O
,	O
generation	O
of	O
profile	O
HMMs	O
,	O
the	O
addition	O
of	O
secondary	O
structure	O
predictions	O
to	O
MSAs	O
,	O
the	O
extraction	O
of	O
alignments	O
from	O
program	O
output	O
,	O
and	O
the	O
generation	O
of	O
customized	O
databases	O
.	O
</s>
<s>
The	O
HMM-HMM	O
alignment	O
algorithm	O
of	O
HHblits	O
and	O
HHsearch	O
was	O
significantly	O
accelerated	O
using	O
vector	B-Device
instructions	I-Device
in	O
version	O
3	O
of	O
the	O
HH-suite	B-Application
.	O
</s>
