Multivariate Analyses of Quality Metrics for Crystal Structures in the PDB Archive

Chenghua Shao, Huanwang Yang, John D. Westbrook, Jasmine Y. Young, Christine Zardecki, Stephen K. Burley

Research output: Contribution to journalArticlepeer-review

16 Scopus citations


Following deployment of an augmented validation system by the Worldwide Protein Data Bank (wwPDB) partnership, the quality of crystal structures entering the PDB has improved. Of significance are improvements in quality measures now prominently displayed in the wwPDB validation report. Comparisons of PDB depositions made before and after introduction of the new reporting system show improvements in quality measures relating to pairwise atom-atom clashes, side-chain torsion angle rotamers, and local agreement between the atomic coordinate structure model and experimental electron density data. These improvements are largely independent of resolution limit and sample molecular weight. No significant improvement in the quality of associated ligands was observed. Principal component analysis revealed that structure quality could be summarized with three measures (Rfree, real-space R factor Z score, and a combined molecular geometry quality metric), which can in turn be reduced to a single overall quality metric readily interpretable by all PDB archive users.

Original languageEnglish (US)
Pages (from-to)458-468
Number of pages11
Issue number3
StatePublished - Mar 7 2017

All Science Journal Classification (ASJC) codes

  • Structural Biology
  • Molecular Biology


  • OneDep
  • PDB
  • RCSB
  • multivariate analysis
  • principal component analysis
  • protein crystal structure
  • structure quality
  • structure validation
  • wwPDB

Fingerprint Dive into the research topics of 'Multivariate Analyses of Quality Metrics for Crystal Structures in the PDB Archive'. Together they form a unique fingerprint.

Cite this