User:Wayne Decatur/1cts to 2cts (citrate synthase) morph methods
From Proteopedia
MORPH
First I removed model numbers from 1cts and 2cts and tried submitting to Yale Morph beta server for multichains. It turns out it just did one chain of a for each and not the dimer. So I kept those files (could be of use for showing the changes in a single chain on the citrate synthase page), renumbered the models from 0 through 7 to 1 through 8 (SO THEY WORK PROPERLY WITH JMOL), and then added back the ligands (citric acid and C to chain 8 by using MagicFit in Swiss-PDB Viewer like described in the 3kg2 morph methods. Turns out since citrc acid in both start and end models, that I should have fitted that part in to all models. However, getting things done for dimer is more important.
To do dimer:
- I removed references to model numbers from 1cts and 2cts.
- Then using the second set of chain A in each PDB file, I changed the chain identifier to B for that part using part of the file in PDB Goodies. It didn't matter that PDB Goodies ignored and left out the heteroatoms because the Morph Server does too. I ended up needing to add the ligands back later to the A and B chain files to do fitting with SWISS-PDB VIEWER to get ligands back, and second set of ligands I just set to chain B by hand since not worth setting up my Python scripts just for a few lines. (For 3cts,5cts, and 6cts biological files, later I did just with word processing using search for A with space in front and behind because didn't need for morph and wanted all hetero atoms. Removed references to model numbers from these 3 files too)
- Then I submitted each of those two produced files subsequently to the Yale Morph Server Beta Server and got the result quickly.
- I downloaded the produced multi-model file by clicking on the Jmol icon in the corner and selecting view movie.pdb.gz from the menu and unzipped the file.
- I renumbered the models from 0 through 11 to 1 to 12 so they'd work with Jmol.
- I added back the citrate and Acetyl-COA to files using the ones I used to generate the files plus the ligands added and labeled (see second step above).
- Gzipped and Uploaded as Image:1ctsTO2cts.pdb.gz
- However, I realized after uploading it and making a scene that I forgot to reduce it to just the alpha carbons first using alphac.exe (2nd edition). So I did that, gzipped, and uploaded as Image:CA1ctsTO2cts.pdb.gz .
|
These two morphs are used on the Citrate Synthase page.
On the Morphs page, I had at one time also used this:.
Testing code for the Morphs page:
NOTE: NEED NUMBER DUE TO JMOL OR PROTEOPEDIA CAUSING ERROR IN CONSOLE IF TRIGGER SELECT COMMAND WITH JUST COLON IN FRONT OF CHAIN DESIGNATION;
-OR-
<p/>
Related to this work, it was suggested a morph like this one keep citrate in all the models (frames) since it is bound at both open and closed state. Because the morph server drops out heteroatoms, adding citrate to each frame, especially each monomer of the dimer, is not trivial and so I hadn't done. However, if I determine what residues are closest to citrate at each state and use the Jmol compare command focused only on those residues to match the open form to the first few frames of the morph and the same with the closed form for the last few frames. I should be able to save a file with the adjust coordinates and move citrate into the full morph of the appropriate frame. I would need to do a pass for each monomer. This should work better than just calculating a linear interpolation for each atom with the distance moved along a straight line divided among the number of frames currently in the animation (using Python) of the citrate atoms between the open and closed form as there may be some twisting and rotating of the citrate along the way that the straight line calculation for each atom wold not accommodate.
- Will be doing this with morph file with all atoms ("1ctsTO2cts.pdb"), because work related to this used that file to show all atoms.
