Solve

Solve manual                                        How to run Solve at BM16    

SOLVE is very easy to use and carries out all the steps involved in phasing your structure from anomalous diffraction data. The program carries out local scaling, automatically interprets the Patterson(s), calculates difference fouriers to find additional sites, and calculates phases.

Manual

The Solve manual can be found here and a list of Solve keywords can be found here.

Running SOLVE

To run Solve all that you need are the scaled data files for each wavelength and an input file (solve.sad or solve.mad).

Before running the program you need to edit the input file to give nformation on the unit cell, space group, file names, values of f" and f' and the number of sites to be found. When you have done this you can run Solve on either bm16process1 or bm16process2 as follows:

%solve < solve.mad > solve.log &

The status of the SOLVE job can be monitored by typing:

%tail -30f solve.status

When the job has finished the output file solve.prt contains a range of useful stats including the completeness of each data set and the size of the anomalous and dispersive differences measured, as well as the peak heights of the theoretical and observed Patterson functions, peaks heights obtained on cross-fouriers, and the overall figure of merit.

SOLVE also produces native electron density maps in both EZD (solve.ezd) and CCP4 formats (solve.ccp4_map), a |Fa|² Patterson map in EZD format (patt_Fa.ezd) and a final mtz file (solve.mtz) containing F, SIGF, PHI, FOM, HLA, HLB, HLC, HLD.

The maps output from Solve can be viewed directly in COOT. A good indication that the structure is solved is when the figure of merit is close to or above 0.5 with few negative peaks on the observed Patterson. Another indication is given by the score associated with the native electron density map; if this is high then the map is likely to be of good quality.

Note: for triclinic, monoclinic and orthorhombic space groups SOLVE is able to distinguish the correct hand of the heavy atom partial structure. This is not the case for other crystal systems, and here SOLVE must be run in both possible enantiomeric space groups.