2.08 *******************
* MISSNG WRITE-UP *
*******************
MISSNG is a program to compare reflections in the main phased data
set with those in other files, and output those reflections absent
in the main file but present in the others. The output file thus
contains candidates for phase extension. Usually the main (phased)
data set is the output from PHASIT and contains all reflections which
survived the cutoffs applied. The additional input files should include
the complete native set, from which the "merged" data files input to
PHASIT were created, but can also include other "phased" files (i.e.
files containing phase probability distribution coefficients). In this
way additional reflections for which native amplitudes are available
(and possibly phase probability distribution coefficients) can be
selected for phase extension via option 3 in BNDRY. The program is
interactive, and prompts for the input and output file names, whether
or not an additional "phased" file is to be included ( perhaps
obtained from a partial structure via PHASIT, SF mode with IHLCF=1 and
ISIGA=0 or 1), a d spacing cutoff, and the parameter file. The output
file contains all additional reflections currently missing from the
PHASIT file out to DCUT resolution for which native amplitudes are
available. If the additional phased file was included, then for
those reflections the phase distribution coefficients are also output
and they will be used during phase combination in BNDRY. For those
reflections without distribution coefficients, the calculated phase
will be output in BNDRY as there is no phase information to combine
with. The output file is generally called "extrfl.d". First the main
phased file is read. Then if an additional phased file is to be used
it is read and the reflections compared with those in the main file.
The additional reflections are then written out along with the
distribution coefficients. The native file is then read and the
indices are transformed to the standard asymmetric unit, and compared
to all reflections previously encountered. Those reflections not
yet utilized will then also be written to the output file, but will
not contain distribution coefficients.
***** FILES *****
The input main phased file should be a PHASIT style output file (long
format, i.e. includes probability distribution coefficients). Only
the indices are used, so that this file may contain either FM*FO and
FO or FO and FC in the amplitude slots.
If an additional phased file is used, it should also be a PHASIT style
output file (long format), but it MUST contain FM*FO and FO in the
amplitude slots as indices, FO and the distribution coefficients are
to be used. It could thus be generated in PHASIT, SF mode with IHLCF=1
and ISIGA=0 or 1, or in PHASIT, phasing mode.
The input complete native file can be one of three types. If the
filename ends with ".MU" or ".mu", then a XENGEN like MULIST is
assumed. Thus the file should contain records with
H, K, L, RES, F, Sig(F), F+, Sig(F+), F-, Sig(F-), Iflag
in format (3I4, 1X, F6.4, 6(1X, F8.2) 1X, I2 ).
The "Iflag" parameter is not used and may be absent.
If the filename ends with ".SCA" or ".sca", then a SCALEPACK
file is assumed. After a variable number of header records
(see the FILE FORMATS section), reflection records follow and
contain
H, K, L, I+, sig(I+), I-, sig(I-)
in format (3I4, 4F8.1)
Note the use of intensities rather than F's. The last two items
in each record may be omitted. If present, they would be used
only if I+ was not measured.
If the file name does not end with ".MU", ".mu", ".SCA" or ".sca"
each record is assumed to contain
H, K, L, F, Sig(F)
and is read in free format, i.e. each item must be separated by at
least one space or a comma. The indices must be INTEGERS and the F and
Sig(F) values REALS.
In all cases the corresponding F values should be identical (or at
least on the same scale) as those input to PHASIT.
The output file is generally named extrfl.d for compatability with the
supplied template command procedures. Its records will contain H, K,
L, FNAT, A_B, C_D in format (3I4, F10.2, 2I12) where the distribution
coefficients are packed two per word in A_B and C_D according to
A_B = ( IFIX(A*100) + 16384 )*32768 + IFIX(B*100) + 16384
C_D = ( IFIX(C*100) + 16384 )*32768 + IFIX(D*100) + 16384
If distribution coefficients are not available the A_B and C_D values
are zero.
This file can be used for phase extension in option 3 of BNDRY.
Note that MISSNG should always be used to prepare the file for phase
extension rather than some other program. This is because MISSNG will
insure that the output indices correspond to the same standard
asymmetric unit as the rest of the program package. If this is not
done, it is possible for redundant (symmetry related) reflections to
creep in to the data set.