Somewhat like here: http://www.topsan.org/@api/deki/files/6421/=coot.png?size=webview Source: http://www.topsan.org/Proteins/JCSG/3etn So if I open this file with the corresponding software I will have the same/similar picture (http://www.topsan.org/@api/deki/files/6421/=coot.png?size=webview)
As well as there must be 3D coordinates of such densities (or somewhat similar; what I mean is if you look at the picture you will see a 3D wire-frame which is an object in 3D space; I want to extract the coordinates of this wireframe-object, representing the corresponding experimental density).
And I am interested in the original density files not anything secondary received from PDB or somewhat.
You can download the structure factor .cif file from the PDB. It contains the experimental data, which is a structure factor amplitude and associated error for all observed Miller indices. If you install some software (e.g. CCP4 or phenix) You can convert that into an .mtz file and calculate a map file. The phases for the map calculation are given by the .pdb model file.
You can also use the EDS (Electron Density Server) which does that for you, and much more.
http://eds.bmc.uu.se/eds/
For some reason EDS doesn't work with all PDB entries. E.g. very old entries don't have structure factors deposited in the PDB.
Be careful with your "wireframe-object" interpretation. The way you set up the map calculation will irreversibly affect your "wireframe-object".
In all practical senses there is no such thing 'original experimental density' for crystallographic structure models. The density is a model of your diffraction data with phases derived from atomic positions (not necessarily all atoms in the protein). So every time you change your atomic coordinates (which you do during model building or refinement) your density changes. As Stefan explained also the way you calculate your density matters a lot.
Your question is not entirely clear. Do you want the coordinates of the nodes in the map? These can be directly recovered from a map file, but keep in mind that these are just sampled from a (sum of) continuous function(s), so a different sampling rate will give you different nodes.
EDS works for most cases, but PDB_REDO has slightly higher coverage of the PDB. You can download maps for the original PDB entry from this URL: http://www.cmbi.ru.nl/pdb_redo/cgi-bin/map0.pl?id=3etn (the last four characters are the PDB identifier). You can also get maps from the PDB_REDO treated model of PDB entries, they are frequently (but obviously not always) better than the original because they are updated using the latest methods of model building, refinement and validation. For 3etn (http://www.cmbi.ru.nl/pdb_redo/et/3etn/index.html), you can see that all model quality indicators are somewhat better. You can get the model and maps form PDB_REDO straight into COOT, but if you want to download files for calculations and such, look here (http://www.cmbi.ru.nl/pdb_redo/content.html) to see what data is available.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Computer Graphics and Animation
- 3D