SECTION SIX: Chemical shift restraints.

After reading NOESY restraints above (if any), read the chemical shift restraints in namelist &shf, or the pseudocontact restraints in namelist &pcshift. Reading this input is triggered by the presence of a SHIFTS or PCSHIFT line in section THREE. In many cases, the user will not prepare this section of the input by hand, but will use the auxiliary programs makeSHF or fantasian to prepare input from simpler files.


Variables in the &shf namelist. (Defaults are only available for shrang, wt, nter, and shcut; you must specify the rest.)

NRING
Number of rings in the system.
NATR(i)
Number of atoms in the i-th ring.
IATR(j,i)
Absolute atom number for the j-th atom of the i-th ring.
NAMR(i)
Eight-character string that labels the i-th ring. The first three characters give the residue name (in caps); the next three characters contain the residue number (right justified); column 7 is blank; column 8 may optionally contain an extra letter to distinguish the two rings of trp, or the 5 or 8 rings of the heme group.
STR(i)
Ring current intensity factor for the i-th ring. Older values are summarized by Cross and Wright, J. Magn. Res. 64:220-231 (1985); more recent empirical parametrizations based on a larger database give improved results (K. Osapay and D.A. Case, J. Am. Chem. Soc.113, 9436-9444 (1991).
NPROT
Number of protons for which penalty functions are to be set up.
IPROT(i)
Absolute atom number of the i-th proton whose shifts are to be evaluated. For equivalent protons, such as methyl groups or rapidly flipping phenylalanine rings, enter all two or three atom numbers in sequence; averaging will be controlled by the wt parameter, described below.
OBS(i)
Observed secondary shift for the i-th proton. This is typically calculated as the observed value minus a random coil reference value.
SHRANG(i)
"Uncertainty" range for the observed shift: if the calculated shift is within SHRANG of the observed shift, then no penalty will be imposed. The default value is zero for all shifts.
WT(i)
Weight to be assigned to this penalty function. Note that this value will be multiplied by the overall weight (if any) given by the SHIFTS command in the assignment of weights (above). Default values are 1.0. For sets of equivalent protons, give a negative weight for all but the last proton in the group; the last proton gets a normal, positive value. The average computed shift of the group will be compared to obs entered for the last proton.
SHCUT
Values of calculated shifts will be printed only if the absolute error between calculated and observed shifts is greater than this value. Default = 0.3 ppm.
NTER
Resiude number of the N-terminus, for protein shift calculations; default = 1.
CTER
Residue number of the C-terminus, for protein shift calculations. Believe it or not, the current code cannot figure this out for itself.


The PCSHIFT module allows the inclusion of pseudocontact shifts as constraints in energy minimization and molecular dynamics calculations on paramagnetic molecules. The pseudocontact shift depends on the magnetic susceptibility anisotropy of the metal ion and on the location of the resonating nucleus with respect to the axes of the magnetic susceptibility tensor. For the nucleus i, it is given by:

where , , and are the direction cosines of the position vector of atom i with respect to the j-th magnetic susceptibility tensor coordinate system, is the distance between the j-th paramagnetic center and the proton i, jax and jrh are the axial and the equatorial anisotropies of the magnetic susceptibility tensor of the j-th paramagnetic center. For a discussion, see: Lucia Banci, Ivano Bertini, Giovanni Gori-Savellini, Andrea Romagnoli, Paola Turano, Mauro Andrea Cremonini, Claudio Luchinat and Harry B. Gray "Pseudocontact shifts as Constraints for Energy minimization and molecular dynamics calculations on solution structures of paramagnetic metalloproteins", Proteins: Structure, Function and Genetics, in press.

The PCSHIFT module to be used needs a namelist file which includes information on the magnetic suscepibility tensor and on the paramagnetic center, and a line of information for each nucleus. This module allows to include more than one paramagnetic center in the calculations. To include pseudocontact shifts as constraints in energy minimization and molecular dynamics calculations the NMROPT flag should be set to 2, and a PCSHIFT=filename statement entered in section THREE.

To perform molecular dynamics calculations it is necessary to eliminate the rotational and traslational degree of freedom about the center of mass (this because during molecular dynamics calculations the relative orientation between the external reference coordinate system and the magnetic anisotropy tensor coordinate system has to be fixed).This option could be obtained with the NTCM, NSCM and NDFMIN flags of SANDER.

Variables in the pcshift namelist.

NPROT
number of pseudocontact shift constraints.
NME
number of paramagnetic centers.
NMPMC
name of the paramagnetic atom
OPTPHI(n)
OPTTET(n)
OPTOMG(n)
OPTA1(n)
OPTA2(n)
the five parameters of the magnetic anisotropy tensor for each paramagnetic center.
OPTKON
force constant for the pseudocontact shift constraints

Following this, there is a line for each nucleus for which the pseudocontact shift information is given has to be added. Each line contains :

IPROT(i)
atom number of the i-th proton whose shift is to be used as constraint.
OBS(i)
observed pseudocontact shift value, in ppm
WT(i)
relative weight
TOLPRO(i)
relative tolerance ix mltpro
MLTPRO(i)
multiplicity of the NMR signal (for example the protons of a methyl group have mltprot(i)=3)

Example. Here is a &pcshf namelist example: a molecule with three paramagnetic centers and 205 pseudocontact shift constraints.

+---------------------------------------------------------------------------------+
| &pcshf									  |
| nprot=205,									  |
| nme=3,									  |
| nmpcm='FE ',									  |
| optphi(1)=-0.315416,								  |
| opttet(1)=0.407499,								  |
| optomg(1)=0.0251676,								  |
| opta1(1)=-71.233,								  |
| opta2(1)=1214.511,								  |
| optphi(2)=0.567127,								  |
| opttet(2)=-0.750526,								  |
| optomg(2)=0.355576,								  |
| opta1(2)=-60.390,								  |
| opta2(2)=377.459,								  |
| optphi(3)=0.451203,								  |
| opttet(3)=-0.0113097,								  |
| optomg(3)=0.334824,								  |
| opta1(3)=-8.657,								  |
| opta2(3)=704.786,								  |
| optkon=30,									  |
| iprot(1)=26, obs(1)=1.140, wt(1)=1.000, tolpro(1)=1.00, mltpro(1)=1,		  |
| iprot(2)=28, obs(2)=2.740, wt(2)=1.000, tolpro(2)=.500, mltpro(2)=1,		  |
| iprot(3)=30, obs(3)=1.170, wt(3)=1.000, tolpro(3)=.500, mltpro(3)=1,		  |
| iprot(4)=32, obs(4)=1.060, wt(4)=1.000, tolpro(4)=.500, mltpro(4)=3,		  |
| iprot(5)=33, obs(5)=1.060, wt(5)=1.000, tolpro(5)=.500, mltpro(5)=3,		  |
| iprot(6)=34, obs(6)=1.060, wt(6)=1.000, tolpro(6)=.500, mltpro(6)=3,		  |
| ...										  |
| ...										  |
| iprot(205)=1215, obs(205)=.730, wt(205)=1.000, tolpro(205)=.500, mltpro(205)=1, |
| &end										  |
+---------------------------------------------------------------------------------+

An mdin file that might go along with this, to perform a maximum of 5000 minimization cycles, starting with 500 cycles of steepest descent. PCSHIFT=./pcs.in redirects the input from the namelist "pcs.in" which contains the pseudocontact shift information.

+-------------------------------------------------------------------+
|								    |
|Example of minimization including pseudocontact shift constraints  |
|								    |
| &cntrl							    |
| ibelly=0,imin=1,ntpr=100,					    |
| ntwx=100,ntwe=100,ioutfm=0,ntr=0,maxcyc=5000,			    |
| ncyc=500,ntmin=1,dx0=0.0001,dxm=1.0,dele=1.0e-07,		    |
| drms=.1,cut=10.,idiel=0, scee=2.0,				    |
| nmropt=2,pencut=0.1, ipnlty=2,				    |
| &end								    |
| &wt type='REST', istep1=0,istep2=1,value1=0.,			    |
|     value2=1.0,  &end						    |
| &wt type='END'  &end						    |
|DISANG=./noe.in						    |
|PCSHIFT=./pcs.in						    |
|LISTOUT=POUT							    |
+-------------------------------------------------------------------+


[Contents] [Previous] [Next]
Updated on January 5, 2000. Comments to case@scripps.edu