Author: Jim Stewart: Department of Chemistry, University of Maryland, College Park, MD 20742, USA
PHACMP allows the comparison of the values of the reflection phases determined in two different ways for the same set of reflections. For example, the phases determined by an MIR calculation can be compared with those determined by a structure factor calculation. Input may either be from two phase sources on the same bdf or from two different bdfs. If two files are used, the reflection set must have been sorted into the same order by the use of the program SORTRF.
PHACMP calculates the mean difference and the weighted mean difference in the phase values, in degrees, for reflections. The estimated standard deviations in the differences are also calculated. The calculations are done as a function of "d" spacing, resolution, in equal volumes of reciprocal space. In addition, counts are made of the values of the phases in both sets as a function of angle. The number of resolution intervals and the range of angle for counts are under user control. Statistics are done only for those reflections which have established phases in both sets. Reflections for which the value of the phase is undetermined in either set are rejected.
All input phases from whatever source are reduced to degrees modulo 360. The differences in phases are defined as acos(cos(p1-p2)) in order to place them all in the range 0° to 180°. Note that the sign of the difference is lost by this method.
The use of this program requires the specification of the two different phase types to be compared. This is accomplished by the use of the phaso1 and phaso2 lines (PHA SOurce) which specify the source of phase information in terms of the types specified to exist in a bdf. The recognized mnemonics are:
The calculated structure factor; items n700 or n801 and n802 in lrrefl:, n is the data set number.
The "best Fourier coefficient" for the "native" or "parent" substance; items 701 and 702 + (m-1)*4, in lrrefl:, m is the alternate solution number.
The "most probable Fourier coefficient" for the "native" or "parent" substance; items 701 and 702 + (m-1)*4, in lrrefl:, m is the alternate solution number.
The calculated phase from density modification methods; item n511 in lrrefl:, n is the data set number.
Direct methods structure factor estimate; items n631 + (m-1)*2, in lrrefl:, where n is the data set number and m is the alternate solution number.
If the two sets are to be input from different bdf's, the second phase set is assumed to reside on a bdf with the extension specified with the ext option (default is PH2). If the second phase set is for a different compound with a different compid, then this must be specified with the option cid on the PHACMP line.
Reads input archive bdf for phase set 1, and optionally phase set 2.
Optionally, reads phase set 2 from a bdf specified with the options cid and ext
title compare FC phase sets for tridi and tridi1
PHACMP list 50 cid tridi2 ext aa1 bin 11.25 shel 10
In this example the the phase set of tridi on the input archive bdf is compared with the phase set on the bdf with extension aa1 for the compound tridi2. Every fiftieth reflection will be listed. The statistics will be made in ten equal volumes of reciprocal space. Counts will be made of the number of reflections having phase values in "bins" from -5.625° to 5.625°, 5.625° to 16.875°, ....,and 343.125° to 354.375°. Counts of differences will be made up through 180°.