<s>
Protein	B-Application
chemical	I-Application
shift	I-Application
prediction	I-Application
is	O
a	O
branch	O
of	O
biomolecular	O
nuclear	O
magnetic	O
resonance	O
spectroscopy	O
that	O
aims	O
to	O
accurately	O
calculate	O
protein	O
chemical	O
shifts	O
from	O
protein	O
coordinates	O
.	O
</s>
<s>
Protein	B-Application
chemical	I-Application
shift	I-Application
prediction	I-Application
was	O
first	O
attempted	O
in	O
the	O
late	O
1960s	O
using	O
semi-empirical	O
methods	O
applied	O
to	O
protein	O
structures	O
solved	O
by	O
X-ray	O
crystallography	O
.	O
</s>
<s>
Since	O
that	O
time	O
protein	B-Application
chemical	I-Application
shift	I-Application
prediction	I-Application
has	O
evolved	O
to	O
employ	O
much	O
more	O
sophisticated	O
approaches	O
including	O
quantum	O
mechanics	O
,	O
machine	O
learning	O
and	O
empirically	O
derived	O
chemical	O
shift	O
hypersurfaces	O
.	O
</s>
<s>
In	O
particular	O
,	O
protein	O
chemical	O
shifts	O
are	O
sensitive	O
not	O
only	O
to	O
substituent	O
or	O
covalent	O
atom	B-Language
effects	O
(	O
such	O
as	O
electronegativity	O
,	O
redox	O
states	O
or	O
ring	O
currents	O
)	O
but	O
they	O
are	O
also	O
sensitive	O
to	O
backbone	O
torsion	O
angles	O
(	O
i.e.	O
</s>
<s>
+Table	O
1	O
:	O
Currently	O
available	O
protein	B-Application
chemical	I-Application
shift	I-Application
prediction	I-Application
programs	O
Name	O
Method	O
Website	O
SHIFTCALCHybrid	O
–	O
empirical	O
chemical	O
shift	O
hypersurfaces	O
in	O
combination	O
with	O
semi-classical	O
calculations	O
https://archive.today/20140324204821/http://nmr.group.shef.ac.uk/NMR/mainpage.htmlSHIFTSHybrid	O
–	O
QM	O
chemical	O
shift	O
hypersurfaces	O
combined	O
with	O
semi-classical	O
calculationshttp://casegroup.rutgers.edu/qshifts/qshifts.htmCheSHIFTQM	O
calculated	O
chemical	O
shift	O
hypersurfaceshttp://cheshift.com/SHIFTXHybrid	O
–	O
empirical	O
chemical	O
shift	O
hypersurfaces	O
in	O
combination	O
with	O
semi-classical	O
calculationshttp://shiftx.wishartlab.comPROSHIFTNeural	O
network	O
model	O
using	O
atomic	O
parameters	O
and	O
sequence	O
informationhttp://www.meilerlab.org/index.php/servers/show	O
?	O
s_id	O
=	O
9SPARTAHybrid	O
-	O
sequence	O
and	O
shift	O
matching	O
to	O
a	O
databases	O
combined	O
with	O
semi-classical	O
calculationshttp://spin.niddk.nih.gov/bax/software/SPARTA/index.htmlSPARTA	O
+Hybrid	O
-	O
sequence	O
and	O
shift	O
matching	O
to	O
a	O
databases	O
combined	O
with	O
semi-classical	O
calculations	O
and	O
artificial	O
neural	O
networkhttp://spin.niddk.nih.gov/bax/software/SPARTA	O
+	O
/CAMSHIFTDistance	O
-based	O
method	O
in	O
combination	O
with	O
parameterized	O
polynomial	O
expansionhttps://web.archive.org/web/20140109151911/http://www-vendruscolo.ch.cam.ac.uk/camshift/camshift.phpSHIFTX2Hybrid	O
–	O
machine	O
learning	O
method	O
using	O
atomic	O
parameters	O
and	O
combination	O
with	O
semi-classical	O
calculations	O
(	O
SHIFTX+	O
)	O
.	O
</s>
<s>
Some	O
methods	O
only	O
calculate	O
or	O
predict	O
chemical	O
shifts	O
for	O
backbone	O
atoms	O
(	O
6	O
atom	B-Language
types	O
)	O
.	O
</s>
<s>
Some	O
calculate	O
chemical	O
shifts	O
for	O
backbone	O
and	O
certain	O
side	O
chain	O
atoms	O
(	O
C	O
and	O
N	O
only	O
)	O
and	O
still	O
others	O
are	O
able	O
to	O
calculate	O
shifts	O
for	O
all	O
atoms	O
(	O
40	O
atom	B-Language
types	O
)	O
.	O
</s>
