Use the data
Find and compare models
On this page
Public entries can be browsed without an account. Start with a protein or dataset of interest, then inspect its associated models and supporting data.
Find an entry
Search the catalog using a PDB identifier, protein name, or other recorded metadata. An entry page brings together its source files, associated models, and related proteins.
Sequence similarity results can help identify related entries. Each match includes sequence identity, coverage, and an alignment. Similar sequences do not necessarily represent the same experimental conditions or biological state.
Inspect a model
Open a model to view its structure and examine the available records:
| View | What to examine |
|---|---|
| Structure | Atomic coordinates in the interactive 3D viewer |
| Files | Model coordinates and associated experimental data |
| Processing history | Recorded software runs and their inputs and outputs |
| Quality metrics | Reported scores, such as R-work, R-free, or correlation coefficients, where available |
Compare models
Begin with models associated with the same entry. Before interpreting differences in their scores, check:
- Input data: Were the models evaluated against identical processed data? Matching file checksums confirm file identity.
- Evaluation method: Were scores calculated using comparable procedures and settings?
- Model coverage: Do the models describe the same chains and residues?
Interpret scores alongside the structures and processing records. A better value for one metric indicates better performance on that measure; it does not establish that a model captures molecular motion more accurately.
Missing files or scores indicate incomplete information, not necessarily poor model quality.
Take the tourOpens a sample entry in a new tab and walks through each part of the page.