<s>
CS23D	B-Application
is	O
a	O
web	O
server	O
to	O
generate	O
3D	O
structural	O
models	O
from	O
NMR	O
chemical	O
shifts	O
.	O
</s>
<s>
CS23D	B-Application
combines	O
maximal	O
fragment	O
assembly	O
with	O
chemical	O
shift	O
threading	O
,	O
de	O
novo	O
structure	O
generation	O
,	O
chemical	O
shift-based	O
torsion	O
angle	O
prediction	O
,	O
and	O
chemical	O
shift	O
refinement	O
.	O
</s>
<s>
CS23D	B-Application
makes	O
use	O
of	O
RefDB	B-Application
and	O
ShiftX	O
.	O
</s>
<s>
CS23D	B-Application
accepts	O
chemical	O
shift	O
files	O
in	O
either	O
SHIFTY	O
or	O
BMRB	O
formats	O
.	O
</s>
<s>
CS23D	B-Application
output	O
consists	O
of	O
a	O
set	O
of	O
10	O
best-score	O
PDB	O
coordinates	O
.	O
</s>
<s>
The	O
overall	O
CS23D	B-Application
score	O
,	O
knowledge-based	O
score	O
,	O
chemical	O
shift	O
score	O
,	O
Ramachandran	O
plot	O
statistics	O
,	O
correlations	O
between	O
the	O
observed	O
and	O
calculated	O
shifts	O
before	O
and	O
after	O
refinement	O
are	O
displayed	O
.	O
</s>
<s>
Homology	O
search	O
:	O
The	O
query	O
sequence	O
is	O
used	O
to	O
find	O
homologous	O
proteins	O
or/and	O
protein	O
fragments	O
in	O
a	O
non-redundant	O
database	O
of	O
PDB	O
sequences	O
and	O
secondary	O
structures	O
of	O
PPT-DB	O
using	O
BLAST	B-Application
.	O
</s>
<s>
Subfragments	O
identified	O
by	O
homology	O
modelling	O
and	O
chemical	O
shift	O
threading	O
steps	O
are	O
assembled	O
into	O
initial	O
3D	O
models	O
using	O
CS23D	B-Application
SFassembler	O
(	O
SubFragment	O
assembler	O
)	O
.	O
</s>
<s>
Model	O
optimization	O
:	O
Model	O
optimization	O
in	O
CS23D	B-Application
is	O
done	O
by	O
a	O
torsion-angle-based	O
minimizer	O
GAfolder	O
(	O
Genetic	O
Algorithm	O
folder	O
)	O
that	O
uses	O
a	O
genetic	O
algorithm	O
to	O
sample	O
conformation	O
space	O
.	O
</s>
<s>
CS23D	B-Application
is	O
a	O
template-based	O
method	O
.	O
</s>
<s>
For	O
a	O
structural	O
solution	O
that	O
is	O
not	O
biased	O
by	O
a	O
template	O
structure	O
or	O
fragment	O
structure	O
,	O
one	O
may	O
want	O
to	O
consider	O
obtaining	O
NOE-based	O
distance	O
restraints	O
(	O
8-10	O
per	O
residue	O
)	O
and	O
using	O
them	O
with	O
the	O
GeNMR	B-Application
program	O
in	O
its	O
ab	O
initio	O
mode	O
.	O
</s>
