Slide 13 of 27
no systematic absences other than those from the Bravais lattice are excluded in processing (e.g. all those absences due to screw axes are included in the results).
It should be stressed at this stage that the penalties are calculated only on the metric symmetry of the cell; it is perfectly possible (and indeed relatively common!) to have a monoclinic cell with b ˜ 90.00º (and thus appear to be orthorhombic), or an orthorhombic cell with a ˜ b (and appear to be tetragonal). It is only possible to be reasonably certain about the true crystal symmetry after merging equivalents and examining the statistics.
Tell Mosflm which solution you have chosen by typing the solution number and space group - prior knowledge has been applied here as the crystal is known to be P64.
Mosflm then refines the solution and applies suitable symmetry - here, forcing a = b, and a = b = 90º, g = 120º. The final SD in spot positions for good spots on an accurately aligned detector should be ~0.5 - 1.0x the pixel size (0.15mm for Mar IP), and around 0.25 - 0.5x the oscillation angle (in this case 1.0º).
The main beam position is refined simultaneously; it should not change by more than a few pixels, and in any case should be known to within 0.5x the minimum spot separation for the expected unit cell, wavelength and distance. Shifts larger than this are commented on in the output!
At each stage the user has the choice of accepting or rejecting the Mosflm’s suggestion. Usually, it is best to accept unless there is contradictory prior knowledge.
Following the autoindexing step, estimate the mosaicity (either by eye/trial and error, or using the built-in function), and run a spot prediction. The coloured boxes should coincide with the observed spots on the image.
Autoindexing can often be improved considerably by using two images 90º apart.