Hi, I agree with Kay, try to fry your native crystals to get the highest overall resolution possible, but go for low resolution if your crystals decay rapidly, particularly when collecting anomalous data. A high overall resolution is always desirable, but during anomalous phasing you can potentially solve good data (no twinning, pseudo-translational NCS symmetry, etc...) even at an anomalous resolution (NOT overall resolution) as low as 5 to 6 Angstroms, though in practice you generally need at least 3 to 4 Angstroms of anomalous resolution for a straight forward solution. Once you have solved the substructure and have initial phases and an acceptable build, you can swap to your native data to build the structure using higher resolution data.
Obviously, there are a number of factors that influence the result: If your crystals are stable, decay very slowly in the X-ray beam and you have a high symmetry space group, you can attempt MAD phasing. However, if your crystals decay rapidly and you have a low symmetry space group you should try SAD phasing. In this case it's best to collect a lot of data (360 to 720 degrees) close to the peak wavelength with high attenuation, fast exposure times and slightly larger oscillations than usual (1 to 2 degrees). My own experience with this has been that larger oscillations work better in this case than fine slicing, which doesn't seem logical. I've seen this with the last three structures I solved by anomalous phasing where 1 degree oscillations gave better anomalous signal that 0.2 degree oscillations with a Pilatus detector at 0.035 second exposures. Maybe it's to do with the fact that you collect more complete data (as in degrees collected) this way compared to fine slicing (where you will cover fewer degrees using the same dose before the crystal decays). You already tried merging several data sets collected from different positions of the same crystal, but if your crystals are isomorphous you can also merge the data from several crystals to improve your SAD phasing as the anomalous resolution often increases with high multiplicity (the overall resolution does not). For this you need to take into account that the heavy metal atoms will be damaged first at the wavelength at which they produce anomalous diffraction, so the best anomalous resolution will be provided by the images at the start of the data even if the overall resolution doesn't seem to decrease. So, check if you get better data by discarding the images to the end of each individual data set before you merge them. Merging and then scaling the data sets together might increase the overall error and not improve the high resolution data, but I have found this will boost the signal of week anomalous scatterers found in the same positions in all the crystals. I hope this helped. Tony ------------------------------------------------------ Dr. Antonio Ariza University of Oxford Sir William Dunn School of Pathology South Parks Road Oxford OX1 3RE
