next up previous contents index
Next: Requirements Up: Setup of the Previous: Setup of the

Syntax

RELAxation { <relaxation-statement> } END
is invoked from the
main level of X-PLOR.
<relaxation-statement>:==

ASSIgn
<selection> <selection> { } adds a single cross-peak intensity or a buildup curve between the two specified selected sets of spins. The number of mixing times read depends on the number of CLASs statements issued before the ASSIgn statement. Error estimates can be specified. The format depends on the ERROr_input statement. For examples, see Section 21.7.
AVERage <real>
specifies an exponent for calculating the average distance to an UNREsolved group of protons (default: 6).
CALIbrate { <calibrate-statement> } END
calculates the calibration factor between the observed and calculated intensities. Prior to the calibration, this command computes the NOE intensities from the current atomic model unless VALUe is specified.
<calibrate-statement>:==
AUTOmatic=<ON,OFF>
sets the calibration factor, if ON, every step during the refinement. At present, this allows the calculation of the gradient only approximately. Therefore, it should be turned OFF during minimization.
GROUp=<CLASs,GROUp,ALL>
for CLASS, calculates the calibration factor separately for each classification; for ALL, one overall calibration factor is used (default: CLASs).
QUALity=<real>
specifies that cross peaks for which the relative error estimate exceeds the specified value are not used (default: 1).
REFErence=<ALL,REFErence>
for ALL, uses all peaks that meet the criterion set with the EXCLude option for the calculation of the calibration factor; for REFErence, uses only those that were flagged as reference peaks during input (default: ALL).
VALUe class <*real*>
sets a value for the calibration factor. Note: if the calibration factor is set by VALUe for a class, it is not recalculated for other classes in the same CALIbrate statement.

CLASsification <class-name>
partitions an intensity table; it applies to all following ASSIgn entries until another CLASS entry is issued. In contrast to the NOE CLASs statement, the RELAxation CLASs statement can be used to switch between classes during the input of cross-peak intensities. Several class statements can be issued without an intervening ASSIgn statement. This allows the input of buildup curves (default: NIL).
CLGR (CLass_GRoup) <*class*> <group>
enters a CLASs into a GROUp of spectra (default: none).
CUTOff
{ <cutoff-statement> } END applies a cutoff to relaxation matrix and/or gradient calculation. When a cutoff is in effect, for each peak on the data list a separate small relaxation matrix is set up, including only spins within a range of the spin pair specified by the cutoff. The spectrum and the gradient are calculated with this small relaxation matrix.
<cutoff-statement>:==
MODE=<NONE,GRADient,ALL>
applies no cutoff, cutoff only for the gradient calculation, or cutoff for both intensity and gradient calculation (default: NONE).
RELAtive_value=<real>
specifies a cutoff relative to the actual distance between the two spins in Å (default: 0).
VALUe=<real>
specifies the actual value of the cutoff, in Å (default: 6).

EEXPonent <real>
is an exponent m of the WELL or PARAbolic energy function (default: 2).
ERROr_input
{ <error-input-statement> } END defines an error input format for the following ASSIgn statement. It remains in effect until another ERROr_input command is issued.
<errorinput-statement>:==
INPUt=<UNIForm,RANGe,PLUSminus>
specifies the input format (0, 1, or 2 error estimates) for the following ASSIgn statements (default: RANGe): for UNIForm, no error estimate is expected by the program, for RANGe, one error estimate, and for PLUSminus, two, where the first is an error estimate on the minus side, the second on the plus side.
MODE=<ABSOlute,RELAtive>
interprets the error estimate in the ASSIgn statements as absolute or relative (default: ABSOlute).
VALUe=<real>
enters the actual value for UNIForm option (default: 0).

GROUp <group-name>
defines a group of several CLASses that are recorded with the same physical conditions, like solvent(HO, DO), field strength, and temperature (default: NIL).
IEXPonent <real>
specifies an exponent n of the calculated and observed intensities in the energy function (default: 1).
IWEIght
{ <iweight-statement> } END applies individual weights to each single term in the energy function. At the moment, the only supported options are weights equal to the inverse of powers of the observed intensity (i.e., ). If the exponent g matches the exponent m of the energy function, the weighting is equivalent to refining relative intensity differences. The minimum weight applied is always 1.
<iweight-statement>:==
CUTOff=<real>
is the maximum weight applied (default: 10000).
EXPOnent=<real>
specifies an exponent g of the observed intensity in the weight function (default: 2).
QUALity=<real>
stipulates that cross peaks for which the relative error estimate exceeds the specified value are not used in determining the maximum weight (default: 1).

MINIntensity <*class*> <real>
sets the minimum detected intensity for each class. The specified value is used as the ``zero" for the intensities (default: smallest number used by X-PLOR).
NREStraints <integer>
is the required parameter that specifies the maximum expected number of intensities. It has to be greater than or equal to the actual number. The parameter is used to assign space dynamically for the intensity list (default: 200).

OCCUpancy <group> <selection> <real>
sets the occupancy for selected spins (for example, 0.9 for exchanging amid protons) (default: 1.0).
OMEGa <*group*> <real>
specifies the spectrometer frequency, in Hz (default: 500*10).
OMIT <group> <selection>
is an obsolete statement that was used in a prerelease version of the relaxation matrix code of X-PLOR. It will be removed in a future version. The statement deletes spins from the relaxation matrix. This is almost the reverse of the SELEct command, but it leaves any existing base of unresolved groups intact (default: none).
POTEntial <WELL,PARAbolic>
provides two alternatives: WELL uses a flat-bottom potential with bounds at error estimates, PARAbolic a (generalized) parabola with exponent EEXP and center at the specified intensity (default: WELL).
PREDict_intensities
<class> { <predict-intensities-statement> } END calculates a spectrum from coordinates with the parameters valid for the specified class. Note that this destroys any existing experimental intensity database. The spectrum can be calculated for parts of a molecular structure.
<PREDict_intensities-statement>:==
CUTOFf <real>
specifies a distance cutoff for cross-peak calculation (default: 10).
CUTON <real>
specifies a distance cuton for cross-peak calculation (default: 0).
DIAGonal <ON,OFF>
prints diagonal peaks (default: OFF).
FORMat <NORMal,LIST>
is an output format; LIST produces an intensity INPUT file (default: NORMal).
FROM <selection>
selects the first part of a molecular structure (default: all NMR-active atoms).
THREshold <real>
specifies the minimum intensity that is printed (default: 0).
TO <selection>
selects the second part of a molecular structure (default: all NMR-active atoms).

PRINt_deviations
{ <print-deviation-statement> } END prints the deviations from observed intensities and calculates an R value. It uses the form of the energy function ( WELL or PARAbolic and the value of IEXP) specified before issuing the statement.
<print-deviation-statement>:==
FORMat <NORMal,LIST>
is an output format; LIST produces a listing in the format of an intensity input file.
SELEct <selection>
calculates the R value with only a subset of the atoms. This allows the calculation of local R values.
THREshold <real>
specifies the minimum deviation that is printed, in the arbitrary units of the input intensities. For a complete list of deviations (including 0), THREshold should be set to a negative number (default: 0).

REFErence
flags intensities read by the following ASSIgn statements as reference peaks for the calculation of the calibration factor (see CALIbrate statement), and remains in effect until another CLASs statement is issued. This flag allows one to use a subset of peaks for the calibration, e.g., peaks that do not depend very much on the conformation of the molecule. The reference peaks are written into a separate file, and the reference flag is set before reading that file.
RESEt
erases the current relaxation database.
SELEct <group> <selection>
selects the ``NMR-active" atoms (default: (HYDRogen) ). At present, only protons can be used in the refinement. The NMR parameters for protons are applied to all selected atoms. This command initializes the array that stores information about unresolved groups (see below). If the command is not issued at the beginning of the relaxation matrix setup, all hydrogens are used by default. Since a hydrogen is defined by its mass in X-PLOR, care should be taken in this case to modify the masses of hydrogens for simulated annealing only after the relaxation matrix parameters have been set up.
SORT_intensities
sorts the intensity list so that peaks with different mixing times belonging to the same spin pair are consecutive on the list. If a cutoff is used, a new relaxation matrix generally has to be set up and diagonalized for each peak on the data list. However, if the only difference between two entries on the data list is the mixing time, the relaxation matrix does not have to be recalculated and diagonalized. Sorting can thus save a considerable amount of CPU time.
TAUCorrelation
<group> { <TAUC-statement> } END specifies the correlation time and the order parameter, depending on the chosen model.
<TAUCorrelation-statement>:==
MODEl <RIGId|LIPAri>
provides two alternatives: RIGId expects no order parameter; LIPAri expects one overall correlation time and one order parameter. This applies to both the ISOTropic and the VECTor case.
ISOTropic
specifies an isotropic correlation time and an order parameter (default: 10 ns, 1).
RESEt
initializes a table.
VECTor <selection> <selection>
specifies a
correlation time (in seconds) and an order parameter for a specified vector.

TAUMix <*class*> <real>
specifies the mixing time , in seconds (default: 0.1 sec).
TOLErance <real>
if not equal to zero, approximates the exact calculation of the energy function and its first derivatives with respect to the atomic coordinates by freezing until atoms have moved by less than TOLErance Å\ from the point at which was last computed. This is particularly useful for molecular dynamics calculations. It should be set to 0 during minimization to avoid inconsistencies in the force field (default: 0).
UPDate
computes the NOE intensities from the current atomic model.

UNREsolved
<group> { <unresolved-statement> } END   replaces protons with unresolved chemical shifts by a single spin. The distance to an unresolved group is calculated as , where the exponent av is set with the AVERage command. A diagonal leakage rate is added to the relaxation matrix for each unresolved group. The main use of this command is for methyl groups.
<unresolved-statement>:==
1-2
treats each methylene group as an unresolved group of protons. A methylene group is recognized by X-PLOR as two hydrogens bound to a heavy atom.
METHyl
treats each methyl group as an unresolved group of protons. A methyl group is recognized by X-PLOR as three hydrogens bound to a heavy atom.
NONE
deletes all previously defined unresolved groups.
SELECT <selection>
treats the selected atoms as an unresolved group of protons.

WEIGht <*class*> <real>
specifies the energy constant (overall weight) (default: 1).
ZLEAkage_rate <*group*>
specifies the overall diagonal leakage rate , in sec (default: 0).



next up previous contents index
Next: Requirements Up: Setup of the Previous: Setup of the



Web Manager
Sat Mar 11 09:37:37 PST 1995