Next: Requirements
Up: Crystallographic Diffraction Data
Previous: Orthogonalization Convention
The xrefin statement is used to read structure factors,
symmetry operators, atomicform factors, unit cell parameters, etc.
It also branches into several other statements that allow
one to manipulate structure factors, refine certain parameters,
carry out translation or rotation searches, compute
solvent masks, or compute
electron density maps.
Manipulations
of structure factors are carried out for the selected
reflections. The selection of reflections is accomplished
by the RESOlution and FWINdow statements. Atoms contributing
to the structure factor are selected through
the SELEction and SCATter statements.
 XREFin { <xrefinstatement> } END
 is invoked
from the main level of XPLOR.
 <xrefinstatement>:==

 A=<real>
 specifies a of unit cell (default: 1.0 Å).
 ALPHa=<real>
 specifies of unit cell
(default: 90).
 B=<real>
 specifies b of unit cell (default: 1.0 Å).
 BETA=<real>
 specifies of unit cell
(default: 90).
 C=<real>
 specifies c of unit cell (default: 1.0 Å).
 DO
 <xrefindostatement> manipulates structure
factors (see Section 12.5).
 EXPAnd
 expands selected reflections to P1; i.e.,
the crystallographic symmetry operators and the Hermitian
symmetry operator (if HERMitian=TRUE) are applied to the
selected reflections. Multiple entries that emerge
during this process are discarded. Multiple application
of EXPAnd
is possible. The second time around, it does not have any
effect unless
new data are added or more symmetry operators are added.
 FFK=<real>
 sets k
in Eq. 12.1 to the specified value.
If FFK is set to 0, k is automatically
determined by Eq. 12.6.
Automatic scaling also declares symbols ($FFK, $TEST_FFK)
that contain the values of k for the working set and the
test set, respectively
(default: 0, i.e., automatic scaling).
 FFT
 { <FFTstatement> } END specifies
parameters for the FFT method.
 FWINdow
 <real> <real> sets amplitude
limits for the selection of reflections. One of the <real>
values is the upper value; the other one is the lower value;
it does not matter whether the upper or the lower
limit comes first (default: FWINdow 0 100000).
 GAMMa=<real>
 specifies of unit cell (default: 90).
 GENErate
 complements the current reflections to yield
a full
asymmetric unit of reflections for the specified resolution range.
If no current reflections
are present, a full asymmetric unit is generated.
The new Fobs are set to 1 except for systematic
absences, in which case Fobs is set to 0.
Fcalc, Fpart are set to 0;
weight, sigma, and FOM are set to 1.
Action is
taken as soon as this statement is issued.
 HERMitian=<logical>
 specifies whether Friedel mates
are identical (HERMitian=TRUE) or different from
each other (HERMitian=FALSE). Thus, for anomalous scattering data, one has to
set HERMitian=FALSE and specify the imaginary
scattering components
for the appropriate atoms (default: TRUE).
 LOOKup=<logical>
 is a flag indicating
whether to use lookup tables
for the direct summation method and the FFT method
(default: TRUE). Turn off the lookup
tables if problems with minimization procedures are
encountered.
 MAP
 { <xrefinmapstatement> } END
computes electron density maps (see
Section 14.1).
 MBINs=<integer>
 specifies the number of bins
for the R value analysis (PRINt R), phase difference
analysis (PRINt PHASe), data completeness
analysis (PRINt COMPleteness), and
computation of normalized structure factors
(Es) (default: 8).
 METHod=
 DIREct  FFT specifies choice of
method to compute (default: FFT).
 NREFlections=<integer>
 is a required parameter
that allocates space for the reflections. It has to be
greater than or equal to the actual number of reflections
(default: 200).
 OPTImize
 BFACtor
{ <xrefinoptimizebfactorstatement> }
END optimizes individual (restrained) isotropic Bfactors (see
Section 13.4).
 OPTImize
 GROUp { <xrefinoptimizegroupstatement> }
END optimizes group Bfactors and/or occupancies (see
Section 13.3).
 OPTImize
 OVERall
{ <xrefinoptimizeoverallstatement> }
END optimizes an overall isotropic or anisotropic
Bfactor (see Section
13.2).
 PRINt COMPleteness
 prints the ratio of the
number of observed reflections
to the number of theoretically observable
reflections (``completeness," printed as
a percentage). The analysis
is carried out as
a function of resolution. The overall
completeness is stored in the symbol $COMPLETENESS.
If the data are partitioned into a test set and a working
set, a
completeness analysis is also carried out
for the test
reflections (see Chapter
15). In this case, the completeness for the
test reflections (TEST=1) is stored
in the symbol $TEST COMPLETENESS, and
the completeness for the working set is stored in
$COMPLETENESS.
This statement
also produces a listing that can be plotted by a
Mathematica script.
 PRINt PHASe
 prints the average phase difference
as a function of resolution for the selected
reflections and stores the overall phase
difference in the symbol $DPHI. Phases differences
are weighted with the WEIGht array (see Section 12.5).
If a figureofmerit
weighted average is required, the
user should issue a ``DO (WEIGHT=MAX(0,FOM))" statement;
this will overwrite the existing weights.
If the data are partitioned into a test set and a working
set, a phase difference analysis for the test
reflections is also carried out (see Chapter
15). The ``free" phase difference
is stored
in the symbol $TEST DPHI. Note that
no prior update of is performed, and thus the
user must issue an UPDAte statement if the
array is undefined or not well defined.
This statement
also produces a listing that can be plotted by a
Mathematica script.
 PRINt R
 prints the R value
as a function of resolution for the selected
reflections and stores the
overall R value in
the symbol $R.
If the data are partitioned into a test set and a working
set, a free
R value analysis is also carried out (see Chapter
15). The free R value is stored
in the symbol $TEST R. Note that
no prior update of is performed. Thus,
the user must update with the UPDAte
statement if the array is undefined or
if an energy calculation (e.g., energy minimization or
molecular dynamics) was carried out previously.
In the latter case, the array is
filled with the values of the derivatives of the
target function. This statement
also produces a listing that can be plotted by a
Mathematica script.
 PRINt TARGet
 prints the value
of (Eq. 12.1). If
the residual or the AB vectorresidual
is chosen, the R value is stored in the symbol $R.
If targets are chosen that contain a correlation coefficient,
its value is stored in the symbol $CORR.
If the data are partitioned into a test and a working
set (see Chapter 15), the corresponding values
for the test set are stored in the symbols $TEST R
and $TEST CORR.
Note that
no prior update of is performed, and thus the
user must issue an UPDAte statement if the
array is undefined or not well defined.
 PRINt WILSon
 makes a Wilson plot (Wilson 1949; Rogers
1965; Main 1975). The overall scale factor and Bfactor
are obtained by a leastsquares fit of
vs. . The result is stored in symbols
$BFACTOR, $SCALE, and $TEST_BFACTOR, $TEST_SCALE for
the working and test sets, respectively. This statement
also produces a listing that can be plotted by a
Mathematica script.
 REDUce
 decreases selected reflections to an
asymmetric unit.
In the case of multiple entries, the first entry is
kept and the duplicates are discarded; i.e., no data
averaging is performed.
 REFLection
 { <xrefinreflectionstatement> } END
initiates reading or
merging of diffraction data (see Section 12.4).
 RESEt
 erases the current xrefin database, i.e., the
atomicform factors, symmetry operators, reflections,
and unit cell parameters.
 RESOLution
 <real> <real> sets resolution limits
in Å for the selection of reflections. One of the
<real> values is the high resolution limit, and the other one
is the low resolution limit; it does not matter whether the high
or the low resolution limit comes first.
One of the limits can be set to the string ``INFInity",
e.g., ``RESOlution INFInity 3".
This statement
will then include the 0,0,0 reflection in all
calculations (default: RESOlution
10. 3. ).
 SCATter
 <selection> <real> <real>
<real> <real> <real>
<real> <real>
<real> <real> [ IMAGinary <real> ]
adds an atomicform factor specification to the xrefin
database. The statement specifies the coefficients (Eq. 12.10)
for the selected atoms
(default: none). An atom
will contribute to the structure factors only if it has
been selected in one SCATter statement and if it
has been selected in the SELEction statement.
Care should be taken not to produce an overlapping
definition of atom selections; e.g., if there is a
ca ion, one
should exclude it from the atomicform factor definition for
carbon atoms. The optional IMAGinary parameter should be used
for anomalous scatterers.
``SCATter RESEt" will erase the existing atomic
form factor entries.
 SEARch
 ROTAtion { <xrefinsearchrotationstatement>
} END is a translation search (see
Section 17.3).
 SEARch
 TRANslation
{ <xrefinsearchtranslationstatement> } END
is a rotation search (see
Section 17.3).
 SELEction=
 <selection> selects atoms that will
be used in the next structure factor calculation (default: (ALL) ).
An atom
will contribute to the structure factor only if it has
been selected in one SCATter statement and
has been selected in the SELEction statement.
The selection remains active until a new SELEction statement
is issued.
 SOLMask
 { <xrefinsolmaskstatement> } END
computes a solvent mask (see Section 12.7).
 SYMMetry=<symmetryoperator>
 adds a new symmetry operator
() to the xrefin database. The
notation is the same as in the International
Tables for Crystallography (Hahn ed. 1987),
e.g.,
Multiple entries specify
the space group. A listing of the symmetry operators
of all crystallographic
space groups is stored in file ``symlib.sym"
in the ``symmetry" directory.
In the case of centered space groups, all symmetry operators
have to be specified; otherwise symmetry images will be missing
for the packing interactions. Normally, this redundant
specification of
symmetry operators represents only a small
increase in CPU time. The option
``SYMMetry RESEt" will erase
the existing symmetry
operators (default: symmetry=(x,y,z)).
 TARGet=
 RESIdual  AB  F1F1  F2F2  E1E1 
E2E2  PACKing specifies choice of the target
function; see Eq. 12.1 (default: RESIdual).
 TOLErance=<real>
 specifies the maximum amount
(in Å) by which
any atomic coordinate can deviate from the position when
was last computed during molecular
dynamics, energy minimization, or energy calculations.
If TOLErance is exceeded, and its
derivatives with respect to atomic parameters
are recomputed (default: 0.5 Å).
 UPDAte
 computes for selected reflections
using the current atomic model and atomicform factors.
 WA=<real>
 specifies the overall weight factor in
Eq. 12.1 (default: 1).
 WP=<real>
 specifies the overall weight factor
for the phase term in Eq. 12.1
(default: 0).
 WRITe
 REFLection { <writereflectionstatement> }
END
writes the specified xrefin properties, such as FOBS or
FCALC, of all selected reflections to a specified output file
(see Section 12.4).
 <FFTstatement>:==

 AVOID=<integer>
 facilitates avoidance of the specified integer as
a factor in the x and y physical dimension
of the 3d electron density matrix in order
to avoid memory conflicts on certain supercomputers
(default: machine dependent, set automatically).
 BASE=<integer>
 specifies the minimum prime allowed in FFT dimensions
(default: machine dependent, set automatically).
 BSCALefactor=<real>
 sets the artificial temperature factor
to minimize aliasing effects (default: 20.0 Å).
 ELIMit=<real>
 defines an atomic ``radius" for the
electron density calculation by specifying the
ratio of the atomicform factor at zero and at the radius
of the atom in natural logarithmic units (default: 7.00).
The electron density outside the radius
is set to zero.
 GRIDsize=<real>
 specifies the grid size relative to
the high resolution
limit (default: 0.33,
which corresponds to 1/3 of the high resolution limit).
 MEMOry=<integer>
 indirectly determines the factorization
of the FFT by specifying the maximum memory
allocation allowed for
the electron density matrix in units of
complex words (default: 500000).
 PRIMe=<integer>
 specifies the maximum prime allowed for FFT
dimensions (default: machine dependent, set automatically).
Next: Requirements
Up: Crystallographic Diffraction Data
Previous: Orthogonalization Convention
Sat Mar 11 09:37:37 PST 1995