Hi Smithy On 16 Jun 2015, at 14:54, Smith Liu wrote:
> Dear All, > > When I do mosaicity estimation by iMOSFLM, it shows, "The mosaicity > estimation has not worked for some reason. Message from Mosflm is - Unable to > estimate mosaicity automatically from this image-determine a value visually. > You should enter an estimated value to replace 0.05". This usually happens when the indexing has not actually worked for some reason - it's very hard to say why without seeing your mosflm.lp file (and sometimes the images). If, for example, you have tried to index using adjacent images from a fine phi-sliced dataset then the autoindexing in Mosflm is less likely to work (unlike XDS where it can really help enormously). If you *really* want to use a lot of images (because you can't get >100 spots from the default 2 images) then use images that are well-separated in phi (e.g. at 0, 15, 30, ...). > First I have input a lot of images, I think " Unable to estimate mosaicity > automatically from this image-determine a value visually" should be " Unable > to estimate mosaicity automatically from theseimages-determine a value > visually". No. The mosaicity is estimated from a single image, no matter how any images you enter. > Secondly, will you please tell me how can I determine a value visually? By "trial and error" - enter values and see if they predict most of the spots on the image (i.e. most of the prediction boxes cover spots, and most of the spots are covered by prediction boxes - it's very likely you will see some boxes where there are no spots, and some spots that are not predicted). It can be a pain in the neck doing this. If you get a lot of low resolution spots that are not predicted, try reducing the "mosaic block size" to a value close to 1 - this has the effect of increasing the mosaicity at low resolution but keeping it almost the same at high resolution (see Nave, C. in J. Synchrotron Rad. (2014). 21, 537-546, "Matching X-ray beam and detector properties to protein crystals of different perfection" for an explanation) > And in which step I can get a reliable mosaicity estimation? After postrefinement. The estimation is there because postrefinement of the mosaic spread works better if it starts with a non-zero positive value than zero. Harry -- Dr Harry Powell, MRC Laboratory of Molecular Biology, Francis Crick Avenue, Cambridge Biomedical Campus, Cambridge CB2 0QH Chairman of International Union of Crystallography Commission on Crystallographic Computing Chairman of European Crystallographic Association SIG9 (Crystallographic Computing)
