TUTORIAL

 

Analysing an active site

 

Building Loops

 

Building a functionnal unit from a monomer

 

Crystal Symmetries

 

Electron Density Maps

 

Energy minimisation

 

Fitting Residues into Electron Density

 

Homology modelling

 

Making Phi/Psi statistics

 

Superposing Proteins

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

home


Last modified
31 March 98
by N.Guex

 

Tutorial : building loops


In this example, we will learn to use the building loop tools.

Step by Step

open three copies of the pdb file 1CRN provided with the tutorial package.

Use the Display menu to bring up the Layer Infos Window.

Hide the second and third copies of the protein.

Now use the "Build Loop" item of the "Build" menu. You will be prompted to pick two residues which will serve as "anchors ones". Pick L18 and A24 (the blue and the green residues). After a while, a window appears, containing a list of possible loops. The currently selected loop appears in red.

clash score: 1
PP:-10.56
FF:200.2
  C-N+    CA-C-N+   C-N+-CA+
----------------------------
  0.04    -1.48    -7.09 
  0.06     0.00     0.00 
 -0.11     0.00     0.00 
 -0.11     0.00     0.00 
  0.36     0.00     0.00 
  0.33     0.00     0.00 
  0.47     0.00     0.00 
  0.36     0.00     0.00 

 

The first column gives the diviation in Å to the ideal closure bond length, while the next two columns give the deviation (in degrees) to the ideal angle closure. As you can see, the first loop is the only one with the values computed. Simply click on a line to see how well fits the next loop. Note that you can also use the up and down arrows keys to browse among the solutions.

To help you find the best loop, a count of the clashes (bad contacts or H-bonds) is displayed at the top of the window. There is also an energy information (computed with a partial implementation of the AMBER force-Field) that appears after the "FF" text, and a mean force potential value (PP) computed from a "Sippl-like" mean force potential [ref. 6].

You can click on either of those lines of the header to sort the results accordingly to this specific criterion. Play a little with the various loops proposed, and then select one that seems good enough.

Now make the second copy of 1crn (the one in the second layer) visible again and compare your best solution with the actual one.

Hide the first copy of 1crn (the one that contains the rebuilt loop) and make the second layer active by clicking on the second protein listed in the Layer Infosd Window.

Now use the "Scan Loop" item of the "Build" menu. As before, you will be prompted to pick two residues which will serve as "anchors ones". Again, pick L18 and A24 (the blue and the green residues). After a while, a window appears, containing a list of possible loops. The currently selected loop appears in red.

The window content is slightly different, and gives you the name of PDB files that contain a suitable loop, the chain identifyer, the starting residue, the sequence of the possible fragment, and the resolution (in Å) at which the structure has been solved. Note that a resolution of 0.0Å means that the structure has been solved by NMR, whereas a resolution of 9.99Å means that the source structure is a model.

PGTPE             clash score:1      bad G->X: 0
. ...  (-3)       bad Phi/Psi:2      bad X->P: 1
HLEHK             PP:-10.46          bad X->P: 2
FF:34753.9        access:0.00        rms:0.00
-------------------------------------------------
2STV     10 HLEHK  2.50
2STV     92 VLNTA  2.50
1PYP    113 NNPID  3.00
4PTI     35 GCRAK  1.50
1EST     61 NQNNG  2.50
5ABP    130 KESAV  1.80
8ADH    140 GTSTF  2.40
8ADH    212 AGAAR  2.40

The header gives you the sequence of the loop you want to build, aligned with the sequence of a fragment selected from a database of folds. The similarity score for the fragment appears under parenthesis and is computed from the PAM200 matrix. In addition to those informations, you have as before, a Force Field score, a Mean Force Potential Score, a clash score, and the number of residues from the source loop that have bad phi/psi angles (in other words, residues that would have phi/psi angles laying out of the allowed zones of the ramachandran plot).

Also, you have to consider that Gly and Pro are somehow special residues that adopt special phi/psi combinations. Gly can accept any kind of combination, whereas Pro have phi angles constrained around -60 degrees. Therefore, the number of "bad" transversions (Gly in the source loop that would not happily become something else in the loop you want to build; or residues that would not happily become a Pro) are also summarized in the header.

As for the previous building loop tool, you can sort the loops by energies or clashes to ease the process of identifying the best loop.

Select a suitable loop, and make the third layer visible to compare with the actual solution.

Remark: After having constructed your loops, an energy minimisation is mandatory. Swiss-PdbViewer does not provide energy miminization facility, but can export scripts to drive external minimisation engines.