Related Experiment Videos
Homo crystallographicus--quo vadis?
Gerard J Kleywegt1, T Alwyn Jones
1Department of Cell and Molecular Biology, Uppsala University, Biomedical Centre, Box 596, SE-751 24 Uppsala, Sweden. gerard@xray.bmc.uu.se
Structure (London, England : 1993)
|April 9, 2002
Summary
This study on macromolecular crystal structures from 1991-2000 found no improvement in quality over time, despite better data fitting. A surprising correlation was noted between data deposition and model accuracy.
Area of Science:
- Structural Biology
- Biophysics
- Crystallography
Background:
- Automated determination and refinement of macromolecular crystal structures necessitate robust quality assessment.
- Evaluating the reliability and quality of deposited macromolecular models is crucial for scientific reproducibility.
Purpose of the Study:
- To analyze trends in the reliability and quality of macromolecular crystal structures deposited between 1991 and 2000.
- To identify factors influencing model quality and data deposition practices in structural biology.
Main Methods:
- Analysis of macromolecular crystal structure data deposited in public databases from 1991 to 2000.
- Assessment of structural quality using Ramachandran analysis.
- Evaluation of model-data fit using free and conventional R values.
Main Results:
- Average resolution of determined structures remained constant over the decade.
- No significant improvement in average structural quality, as measured by Ramachandran analysis, was observed over time.
- A decrease in the discrepancy between free and conventional R values indicated an improving fit of models to diffraction data.
- A surprising correlation was found between the propensity of crystallographers to deposit experimental data and the free R values of their models.
Conclusions:
- Despite advancements in automation, the overall quality of deposited macromolecular crystal structures did not improve between 1991 and 2000.
- While model-data fit improved, indicating better refinement practices, this did not translate to enhanced structural quality metrics.
- The observed correlation between data deposition and free R values warrants further investigation into factors influencing data sharing and model reliability in structural biology.