Quality Assessment of a Structure: Homoloy Model

For assessing the quality of a homology model, the Swiss Model server provides several parameters with QMEAN in the center. Clicking the question mark in the modelling window (after making a model) will tell us that QMEAN4 scoring function (the original paper Benkert et al. 2008, and a more recent paper Benkert et al., 2012) is a linear combination of four structural descriptors:
-The local geometry is analyzed by a torsion angle potential over three consecutive amino acids.
-Two distance-dependent interaction potentials are used to assess long-range interactions: First, at a residue-level it is based on C-beta atoms only, at the second level an all-atom potential is used.
-A solvation energy is calculated to investigate the burial status (accessibility to water) of the residues.
When using QMEAN as a quality measure to guide the modelling pross, in the original paper it is stated that "QMEAN shows a statistically significant improvement over nearly all quality measures describing the ability of the scoring function to identify the native structure and to discriminate good from bad models".

In the image below we can see the difference between a the two models we made in the first tutorial, one based on the 2.1 Å X-ray structure (left, QMEAN -2.65) and the second based on the 7.5 Å coordinate file from an EM reconstruction (middle, QMEAN -6.48). On the right with QMEAN of -0.86 is PDB entry 1g8p (the experimental structure):

My Image
My Image
My Image

For a comprehensive model assessment we can also click on the “Structure assessment” button shown under the graphics image of the structure in the modelling window:

My Image

This will open a new page with a detailed summary of possible model problems: Bad angles, % of residues in favored regions of the Ramachandran plot, number of outliers, and in addition per-residue QMEAN calculations. This is a wonderful tool, since it allows a detailed comparison of different models. It can also be accessed directly from the menu in “My work space”, which means we can compare a model built at Swiss Model with models built as other servers using the same method and the assessment parameters. Below you can see a comparison of the models generated in the modelling tutorial with the original 1g8p PDB entry:

My Image

Built with 7.5 Å template

My Image

Built with 2.1 Å template

My Image

PDB entry 1g8p

My Image
My Image
My Image

The comparison shows that the quality of the experimental structure is much higher. This can also be seen in the comparison of the Ramachandran plots above. The Ramachandran plots also clearly show that the quality of the experimental structure is better. The experimental structure also has a much smaller number of bad angles, Ramachandran outliers, etc. (not shown). This procedure can be applied for the comparison of models created at different servers.

This concludes the modeling section of the practical applications part.