User:Wayne Decatur/3ewsSLASH3g0h Morph methods
From Proteopedia
Morph from 3ews to 3g0h structure
|
.
Took the two files and submitted them. (Did it with 8 total models.)
The structures did have nucleic acids; however, I knew from past experience the beta server handled them (even though it sitll says it will be updated to handle it) so I used it. (This time though the result just had chain A since I didn't have the nucleic in both structures.)
I got the e-mail and followed the link to a Jmol animation of the morph. I right clicked on the Jmol frank and click the top entry in the menu and then the bottom spot in the menu that came up to download the structures in the morph.
By the way, the structure that came back from the server didn't have secondary structure when I looked at it in PyMol so I chose from the action menu on the right 'assign sec. str.' and it assigned it for all models and was very close to the crystal secondary structure. Since it seemed to be just for viewing and I didn't save this though, it didn't affect the morph file.
Since the morph file that came back just had the protein from chain A from each endpoint structure, I added back the other parts (ligands and RNA, etc.) from 3ews to model 1 and the other parts of 3g0h to Model 8.
However, while that worked for model 8, the ligand of model 1 was off to the side because it needed translating to where model 1 was in the morph which is not the same as 3ews. To move, I copied the 3ews part from the morph and opened it in Swiss-PDB Viewer. Then with MagicFit allowed in 'Loading protein' under preference, I loaded the actual 3ews pdb file. Then I saved the new layer and extracted the information for ADP and placed that as the ADP heteratom information in the Morph file for model 1. After doing that, I realized the first few residues of 3ews were missing too because they were absent in 3g0h and so I extracted the information fro them as well from the file and added them to the model 1 of the morph.
Uploaded to Proteopedia Image:3ews3g0hmorphBETTER.pdb.gz.
loaded '3ews3g0hmorphBETTER.pdb.gz' in Scene Authoring Tools.
- 50-90 = N-terminal and is forestgreen
- 91-298 = Domain 1 and is khaki gold
- 304-480 = Domain 2 and is crimson REALLY USED 297-480 crimson firebrick darkred
- chain e is darkseagreen
I'd have to check but I think where I am using structure TURN; below, I probably wanted structure NONE;. The representation is the same except for color and so it worked for the purposes in which I used it here. But it wasn't probably the best way to do it. See Secondary structure and here and here and here for more information.
When the animation played in Jmol the models corresponding to 3ews and 3g0h had regions that should be turns that were showing as helices and beta sheets and so I used structure command to fix.
select 67-90:a and 1.1;
structure TURN;
select 78-80:a and 1.1;
structure SHEET;
select 94-98:a and 1.1;
structure HELIX;
select 1.1;
cartoon on;
select 297-304:a and 1.1;
structure TURN;
select 428-436:a and 1.1;
structure TURN;
select 466-470:a and 1.1;
structure TURN;
select 1.1;
cartoon on;
select 72-74:a and 1.8;
structure TURN;
select 94-98:a and 1.8;
structure HELIX;
select 128-132:a and 1.8;
structure TURN;
select 128-132:a and 1.8;
structure TURN;
select 297-299:a and 1.8;
structure TURN;
select 231-236:a and 1.8;
structure HELIX;
select 460-468:a and 1.8;
structure TURN;
select 464-466:a and 1.8;
structure SHEET;
select 1.8;
cartoon on;
select 297-304:a and 1.2;
structure TURN;
select 1.2;
cartoon on;
select 297-304:a and 1.3;
structure TURN;
select 1.3;
cartoon on;
select 297-304:a and 1.4;
structure TURN;
select 1.4;
cartoon on;
select 297-299:a;
structure TURN;
select protein;
cartoon on;
Later for showing the N-terminal helix stretched from the RNA binding site to the ATP binding site, I wanted just model 1.1 and 1.8 to show up at the same time so I told it to display all and then set display to just those two: model all; display (*/1.1,*/1.8); model all;
- This was described under frame at the Jmol Interactive script documentation -- "Same as the animation frame command. See also model. Note that you can show specific pairs or sets of frames or models by using frame all followed by display (*/n,*/m,*/p), where n, m, and p are frame numbers." HOWEVER, WHILE THIS LOOKED GOOD ON THE SCREEN IN PROTEOPEDIA WHEN I PRESSED SAVED SCENE, IT ONLY KEPT SHOWING ONE MODEL WHEN LOADED FROM A GREEN LINK OR FROM LOAD SCENE TAB, which is odd because same thing seems to work for ribosome but that is only two models and both next to each other and all just on and not a subset. (From a later test, it seems it would have worked with a template, but not clear how to load correct molecule if not already loaded - which could be eliminated as a problem if use a jmol window only for that set of models.) So I will make a file like that of the two. And edited the commands fixing secondary structure to match.
select 67-90:a and 1.1; structure TURN; select 78-80:a and 1.1; structure SHEET; select 94-98:a and 1.1; structure HELIX; select 1.1; cartoon on; select 297-304:a and 1.1; structure TURN; select 428-436:a and 1.1; structure TURN; select 466-470:a and 1.1; structure TURN; select 1.1; cartoon on; select 72-74:a and 1.2; structure TURN; select 94-98:a and 1.2; structure HELIX; select 128-132:a and 1.2; structure TURN; select 128-132:a and 1.2; structure TURN; select 297-299:a and 1.2; structure TURN; select 231-236:a and 1.2; structure HELIX; select 460-468:a and 1.2; structure TURN; select 464-466:a and 1.2; structure SHEET; select 1.2; cartoon on;
[Control the animation with the 'animation' submenu on the menu that comes up if you click on the Jmol frank in the bottom right corner. Also, if the animation seems to be stuck, scroll in the bar on the right of your browser.]
The .
AS I SAW WITH 3fpn's file today: ENTERING 'model all;' INTO THE CONSOLE DOESNT WORK ON THESE FILES FROM YALE SERVER?!?! I assume it is the sever and not the number of models in the file because the 'model all' approach works with the ribosome files I made as well as LepA's files. uses ---->3ews3g0hmorphBETTER.pdb.gz at Image:3ews3g0hmorphBETTER.pdb.gz. THE SCENE FAILS TO SHOW MORE THAN ONE MODEL!!!
Maybe it is the 'model 0' that Yale Moprh server puts because I was able to use 3ews3g0hJUSTSTARTEND.pdb which I simpled cut out in between models from first and last and made first model 1. Otherwise the formatting is like Yale Morph Server outputs. uses ---->3ews3g0hJUSTSTARTEND.pdb.gz at Image:3ews3g0hJUSTSTARTEND.pdb.gz. YES IT MUST BE 'model 0' stuff or number of models because WORKS!!!!
Paper on the structure
- Collins R, Karlberg T, Lehtio L, Schutz P, van den Berg S, Dahlgren LG, Hammarstrom M, Weigelt J, Schuler H. The DEXD/H-box RNA helicase DDX19 is regulated by an {alpha}-helical switch. J Biol Chem. 2009 Apr 17;284(16):10296-300. Epub 2009 Feb 25. PMID:19244245 doi:10.1074/jbc.C900018200
