0%

​

Share this page with:

Model quality

Assessing model quality is critical when interpreting structural annotations and data. High- and low-confidence regions can directly affect how you understand ligands, domains, or functional annotations. To learn more about why model quality matters, see our new course: Analysing and evaluating macromolecular models.

For NMR, EM, and X-ray entries, the top half of the Model Quality tab looks very similar:

  • Model quality sliders summarise the distribution of validation scores
  • PDBe-REDO sliders (for X-ray entries) show improvements compared to the original model
  • The Mol* viewer on the left displays the structure coloured by quality scores, helping you visually identify regions of concern.

You can filter by issue type (e.g. geometry outliers, clashes, poor fit to data) to highlight specific problems in the 3D view

Overview of the ‘Model Quality’ tab for PDB entries from different experimental methods. For each entry, the interface provides a quick summary of validation scores via percentile sliders, which compare the model against the PDB archive. The accompanying Mol* viewer displays the structure colored by the number of validation criteria outliers, allowing for the rapid visual identification of potentially problematic regions in the model.
Specific issues can be highlighted on the structure. Here we see RSRZ outliers in red.

To understand more about these issues, please refer to our Analysing and evaluating macromolecular models course.

Scrolling down reveals experiment and data quality details:

  • X-ray entries include the data you would normally find in Table 1 of a publication (e.g. resolution, R-factors, completeness)
  • NMR and EM entries do not yet display equivalent details, but these will be added in future updates

Model Quality in 1D Sequence and 2D topology views

Model quality is also integrated into the Macromolecules tabs:

  • In the 1D sequence view, each residue is coloured by its build quality
  • Build quality comes from the wwPDB validation report
  • Colour scheme:
    • Grey= no geometric outliers
    • Yellow / Orange / Red = increasing number of outliers

The 2D topology diagram for each chain is also coloured according to build quality, giving you another way to spot problematic regions.

The 2D layout is coloured according to the model quality.