BLEND is mainly applicable to cases with many more than 3 data sets. So I would say it does notadd anything useful to what already suggested by Matthias.
I have had positive results when scaling and merging several (unmerged) anisotropic data togetherwith POINTLESS / AIMLESS. The main reason for this was that the main directions of anisotropy were not exactly matching and some reflections intensity became stronger. As a result, resolution (as described by CC1/2) increased along that specific direction. I have never tried a similar approach with XSCALE, but I would be interested in learning from peoplewho have tried it. J Dr James Foadi PhD Membrane Protein Laboratory Diamond Light Source Ltd. Diamond House Harwell Science and Innovation Campus Didcot Oxfordshire OX11 0DE United Kingdom office email: james.fo...@diamond.ac.uk alternative email: j.fo...@imperial.ac.uk personal web page: http://www.jfoadi.me.uk On Wednesday, 12 November 2014, 23:18, Antony Oliver <antony.oli...@sussex.ac.uk> wrote: Would the CCP4 program BLEND be a suitable initial option? And then the anisotropic server? Tony. --- sent from my mobile account --- On 12 Nov 2014, at 21:29, Robert Keenan <bkee...@uchicago.edu> wrote: I have three datasets of varying quality collected from different regions of a single crystal. In each case, the data are anisotropic (from Aimless using CC1/2>0.5): Dataset 1: 3.5, 3.5 5.5 ADataset 2: 4.2, 4.3, 4.8 ADataset 3: 3.7, 3.9, 4.4 A I initially took a simple-minded approach and processed each dataset at the appropriate resolution limit (set 1: 3.5A; set 2: 4.2A; set 3: 3.7A) using XDS/XSCALE as implemented in xia2. The resulting merged dataset seems fine (albeit with ugly stats in the high-res bins because of the anisotropy), and it allowed me to solve the structure (Rfree/Rcryst ~31/26). Now I am wondering if I can improve the maps further by first applying an ellipsoidal truncation (using the UCLA diffraction anisotropy server) and then scaling/merging the three datasets together. However, it seems that the UCLA anisotropy server only allows input of one dataset at a time, and it outputs merged amplitudes. Is there some way to obtain the elliptically truncated but unmerged data for each of the three datasets? More generally, are there preferred strategies for dealing with strongly anisotropic data? Bob ---------------- Robert Keenan Associate Professor Dept. of Biochemistry & Molecular Biology GCIS W238 University of Chicago 929 East 57th Street Chicago, IL 60637 (o) 773.834.2292 (f) 773.834.5416 (e) bkee...@uchicago.edu http://keenanlab.bsd.uchicago.edu
