49
edits
SHELX C/D/E: Difference between revisions
no edit summary
No edit summary |
No edit summary |
||
| Line 38: | Line 38: | ||
shelxe xx.fcf yy [reads xx.fcf, yy.hkl, yy.res] | shelxe xx.fcf yy [reads xx.fcf, yy.hkl, yy.res] | ||
xx.hkl contains native data, yy.hkl contains | xx.hkl contains native data, yy.hkl contains F<sub>A</sub> and α and should have been created using SHELXC or XPREP. xx.phi has .phs format (h,k,l,F,fom,φ in free format) and can be made by renaming a .phs output file from SHELXE, but only the starting phases are read from it; if a .phi file is read, the cell and symmetry are read from xx.ins and the native F-values are read from xx.hkl. xx.fcf (from a SHELXL structure refinement) provides cell, symmetry and starting phases. The output phases are written to xx.phs, the log file is written to xx.lst and, if -b is set, improved substructure phases are output to xx.pha and revised heavy atoms to xx.hat.<br> | ||
The first two modes provide density modification starting from atoms or phases, the third and fourth modes are for phase extension, the fifth is an inverse cross-Fourier for finding heavy atoms for a second derivative (yy) with the same origin as the first (xx), and the last mode is useful to confirm the heavy atom substructure from the final refined phases. This is useful as a post-mortem if SAD or MAD phasing fails but the structure could be solved by other means. For these last two modes, the phases for the inverse Fourier are (phi(nat) – alpha), where phi(nat) may be refined (-m etc.) and alpha is taken from yy.hkl. A few cycles of phase refinement may reduce the noise in such maps by improving the weights.<br> | The first two modes provide density modification starting from atoms or phases, the third and fourth modes are for phase extension, the fifth is an inverse cross-Fourier for finding heavy atoms for a second derivative (yy) with the same origin as the first (xx), and the last mode is useful to confirm the heavy atom substructure from the final refined phases. This is useful as a post-mortem if SAD or MAD phasing fails but the structure could be solved by other means. For these last two modes, the phases for the inverse Fourier are (phi(nat) – alpha), where phi(nat) may be refined (-m etc.) and alpha is taken from yy.hkl. A few cycles of phase refinement may reduce the noise in such maps by improving the weights.<br> | ||