Model Evaluation: 4g83

Results for the evaluation of the complex based on PDB Code: 4g83

Background

PDB Code 4g83 belongs to the protein complex of Tumoral Protein p73, which is part of the family of Tumoral Proteins p53.

This protein is involved in the cell cycle regulation, and apoptosis induction, so a binding to DNA is required. This complex is formed by a tetramer, result of the union of two homodimers (Two protein chains and two DNA strands).

Results are shown below. Computational models are represented in brown color, while the crystallographic structure is shown in cyan.

Project files

607KB

4g83.pdb

4g83 Crystal Structure

277KB

4g83_superimposition.pdb

Superimposition-based model

286KB

4g83_modeller.pdb

Modeller-based model

Reconstruction of the Model by Structural Superimposition.

This structure is a bit more difficult than the structure 1o4x because the complex contains an homodimer. That means that we only have one template for two chains in different position.

Side to side comparison of the Crystal structure obtained from PDB (Code: 4g83) and the Superimposition model.

Animation showing the structural alignment between the native structure and the model.

As we can see, the structure superimposition approach performed perfectly, and the model is practically identical to the crystallized structure, the structural superimposition of the two structures was performed in Chimera USCF, by using the algorithm "MatchMaker" and returns a RMSD of 0.000 amstrongs.

Reconstruction of the Complex by using Modeller.

When using Modeller, and giving only the pairwise structures of the interactions from the original complex, the algorithm did not perform as expected. Below you can see the results.

Crystallographic structure against Modeller-based Algorithm results

Animation showing the structural alignment between the crystal structure and the modeller-based approach model

We can see that for the case of an homodimer, the Modeller-based approach algorithm fails at reconstructing the complex.

The structural superimposition of the two structures was performed in Chimera USCF, by using the algorithm "MatchMaker" and returns a RMSD of 0.873 amstrongs for 162 pruned atoms. If we look at all the residues, the RMSD is 10.06 amstrong, meaning that the Modeller-based algorithm did not perform correctly for the creation of an homodimer complex.