[Kinemage Web Site]

Structure Validation Workshop

Using All-Atom Contact Analysis for Model Diagnosis & Repair

[MolProbity Server]

Part 2: Repair of problem sidechains using the interactive KiNG/Probe tools

This part 2 of the All-atoms Contacts tutorial presents two practicals in rebuilding two problematic side chains. The first is Thr 159 found in 2SIM, for which you prepared the contact-dot kinemage in Part 1 of the tutorial. This repair is done without reference to an electron density map (structure factors were not deposited) and so, changes to the model are limited to placing atoms in the same general area as the beginning model. The second practical is of a model in refinement/rebuilding with electron density and so, changes to the model can be greater with another round of refinement to check the proposed rebuilds.

A: 2SIM Thr 159 - View in KiNG

  1. Get the programs: If you have not done so already, download and install the program KiNG (Probe will be included in the KiNG package) for your system, from the site at kinemage.biochem.duke.edu/software/. Instructions for installation are available from a link on the KiNG download page.
  2. To work with the file, 2SIMH-contact.kin, you just downloaded from MolProbity at the end of part 1:
  3. Enlarge the graphics window, if desired, but leave room for later small working windows.
  4. Choose menu item Edit>Find point and enter '159 cb' into the box , with pickcenter on. Zoom in until only half a dozen residues show, and narrow the clipping slab to see clearly. (zoom with right-click + vertical mouse drag; slab with right-click + horizontal drag; translate in Z with middle click + horizontal drag in the upper region of the display) Turn off all the Cbdev balls (the c-beta button), now that you know this one is big, and make sure H's are turned on. The 3 bad clashes surround the Thr methyl, bumping a backbone NH and 2 waters. Rotate in 3D to look down the Ca-Cb bond, to see that chi1 is eclipsed. Now you will repair this sidechain by finding a good rotamer in ideal geometry that has good contacts while occupying the same region of space as the original model.

  --  click on a thumbnail to see a larger screen capture in a separate window  --  

Set Up Rotamers and Torsions

  1. Turn off the 'dots' button to remove the original contact dots.
  2. The rotamer library is invoked in KiNG via the menu item: Tools >Structural biology>Sidechain rotator. When invoking this tool, another tool, the Model manager, will prompt you to open the PDB file for your model. Answer that query (2SIMH.pdb) Move on by ID'ing the residue of interest (middle click on any side-chain atom of Thr159). A pop-up window will appear allowing sidechain dihedral adjustments and rotamer selection.

  --  click on a thumbnail to see a larger screen capture in a separate window  --  

Evaluate Contacts, and Choose

  1. To update contact dots automatically as you adjust the sidechain, check the 'Probe dots' box in the Model manager window. Turn off the old 'dots' button, if you have not already done so in KiNG's side panel.
  2. The dialog box proposes a command line for Probe: remove the words "not water" so you can evaluate the H-bonds to the waters, press either the "Enter" or "Return" keys to accept, and Probe will return the contact analysis for the present conformation, updated as you change it.
  3. Try all 3 rotamers (p, t, or m for plus, trans, or minus chi1), looking for one with good contacts that occupies about the same space as the original model. The p rotamer fits quite well, but its Og1 does not H-bond with the backbone NH of Ile161. Try changing chi1 to form that H-bond, while adjusting/rotating the OH (2nd torsion) to maintain its water H-bond. Forming the sc-mc H-bond causes a clash of one of the methyl hydrogens with the carbonyl oxygen of Val157. KiNG does not do methyl rotations. So, the trick here is to balance forming the Thr159 Og1 to Ile161 H--N H-bond against the nascent clash -- while imagining a methyl rotation of up to 10 degrees from stagger to relieve the clash.
  4. The choice between the central-value p rotamer and this modified conformation should be determined by their fit to the electron density, but unfortunately the structure factors were not deposited. However, both the library rotamer and its modified conformation are in the same local minimum and both are enormously better than the original.
  5. There is no need to save the changes for this lesson. However, When you have resolved the clashes to your satisfaction, you can accept your new Thr159 model: 'Release' the checked out residue from both tools and accept. If desired, save the modified pdb file in the Model Manager.
  6. Quit from KiNG.

  --  click on a thumbnail to see a larger screen capture in a separate window  --  

B: Cth-833 Arg B 54 - View in KiNG

  1. Since this is a new SECSG structure not yet available from the PDB, make sure you have the files cth833_0410011335_refmac1_qcH.pdb and cth833_refmac1_qcH-contact_map.kin in your working directory, and open the kinemage file in KiNG.
  2. Choose the view for 'b 54 Arg-Asp', which showed up in MolProbity's clashlist. Turn on the 2Fo-Fc map (here preloaded for just the sections you'll need, but normally accessed thru the structural biology function in KiNG). Move around, to see that both sidechains fit their density pretty well but make 2 serious clashes. Turn on H-bonds and small overlaps, to see that the Arg H-bonding is OK to the water but has poor geometry to the Asp. Now let's see if you can do better.

  --  click on a thumbnail to see a larger screen capture in a separate window  --  

Choose & Optimize an Arg B 54 Rotamer

  1. As you did in A above, turn off the 'dots' button, choose menu item: Tools >Structural biology>Sidechain rotator. When invoking this tool, another tool, the Model manager, will prompt you to open the PDB file for your model. Answer that query (cth833_0410011335_refmac1_qcH.pdb); because the PDB file contains alternate conformations, KiNG will also ask which set of conformations to use (in this case answer 'B'). Move on by ID'ing the residue of interest (middle click on any side-chain atom of Arg 54). Check the 'Probe dots' box in the Model manager window, and again delete 'not water' from the probe command line, accepting that command change by pressing the "enter" or "return" keys.
  2. You want to see if there is another rotamer capable of fitting the density but with better interactions. To narrow down the 27 choices for Arg, look down the Calpha-Cbeta vector to see if the chi1 choice is clear; in this case the density shows that Cg must lie opposite the backbone NH, which means a trans chi1. Try each one of the listed rotamers that starts with t, making a note of all that position the guanidinium near the right place and in nearly the right plane (there should be 2, one of which is similar to the original conformation and the other has the guanidinium 'flipped' over. Choose that flipped rotamer, and use the chi rotations to optimize its fit in the density and its H-bonds while minimizing any clashes. Use Views 5 & 6 to check fit from all directions. Are you convinced that this conformation is definitely better than the original?
  3. For one more cross-check, go to the 'a 54 Arg-Asp' view and turn on the original dots, to see how this residue was fit in the other subunit.
  4. Quit from KiNG.

  --  click on a thumbnail to see a larger screen capture in a separate window  --  

http://kinemage.biochem.duke.edu/teaching/aca2005/
Jane & Dave Richardson
previous: Part 1 next: Part 3