[phenixbb] How to generate LLG map
rjr27 at cam.ac.uk
Wed Mar 25 03:42:01 PDT 2020
If you have a refined model and you want to use the anomalous signal in your data to identify which atoms are anomalous scatterers, the best way is to use the MR-SAD approach and search for purely imaginary scatterers (element type “AX”). The way this works is that the atoms in your refined model are treated as purely real scatterers and then Phaser uses log-likelihood-gradient maps to find places where adding just the anomalous scattering component would improve the fit to the data. If you allow sites to be added and refined iteratively, at each cycle Phaser will compute a log-likelihood-gradient map to find new sites with significant signal. The last log-likelihood-gradient map will have no peaks above the limit you set (by default, Phaser looks for peaks above 6 times the rms deviation of the log-likelihood-gradient map), but the sites will tell you where there are likely to be anomalous scatterers. This can be done in the Phenix GUI by choosing Phaser-EP under the Experimental phasing tab, selecting MR-SAD phasing for Phaser mode, providing your data, wavelength, and refined structure under the “Input and general options” tab. Give 100% as the sequence identity for the partial model. Under the “ASU contents and sites” tab, specify the composition by providing the sequence and enter AX in the “Scattering types” box, then run.
The AX atom type is defined as something with no real scattering (f0+f’ = 0) and an imaginary scattering component equivalent to 1 electron (f” = 1). After refinement of the occupancies, you can interpret the occupancy of an AX atom as the product of the actual occupancy of the atom and the f” for the element. So, with good data collected using a wavelength around 1.7A, you would expect that an AX atom corresponding to a fully-occupied S atom would have an occupancy around 0.7 (i.e. 1 times 0.67, with some random error from measuring and scaling the data).
If you want to see the initial log-likelihood-gradient map without completing the anomalous substructure, you have to do two things after selecting “Other settings” under the “Input and general options” tab, making sure that you’ve set the User level to at least Intermediate. First, check the box labelled “Write out LLG maps”; otherwise, the map coefficients for the log-likelihood-gradient map will not be written. Second, change the “Maximum number of cycles of LLG structure completion” to 0, so that you see the initial map before any sites have been defined. The map coefficients will appear in a file with a name ending .llgmaps.mtz. Note that the initial map probably will not show everything that could be found in the iterative procedure, because the log-likelihood-gradient maps become more sensitive as the model improves when more sites are added and refined.
Randy J. Read
Department of Haematology, University of Cambridge
Cambridge Institute for Medical Research Tel: +44 1223 336500
The Keith Peters Building Fax: +44 1223 336827
Hills Road E-mail: rjr27 at cam.ac.uk
Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk
> On 24 Mar 2020, at 19:50, afshan begum <afshan490 at yahoo.com> wrote:
> Dear expert,
> I have collect the data with the bound ligand at 1.4 Å . The ligand contain the sulphate molecule in its ,how could we generate the anamolous signal of sulphate ion to confirm its presence in the structure, actually there are very clear density of the full ligand molecules except for the SO3 ion . As I did it before via phenix but I completely forget it was 4 years ago I used to generate the LLG map for the presence of chloride ion .
> Many thanks
> Sent from my iPhone
> phenixbb mailing list
> phenixbb at phenix-online.org
> Unsubscribe: phenixbb-leave at phenix-online.org
More information about the phenixbb