Quality of protein crystal structures.

Area of Science:

• Structural biology
• Biochemistry
• Genomics

Background:

• The genomics era has led to numerous accessible databases for scientific interpretation.
• Investigators often assume data from scientific databases, including the Protein Data Bank (PDB), are error-free.
• However, experimental data, like PDB coordinates, can contain errors and subjective interpretations.

Purpose of the Study:

• To explore determinants of protein structure quality metrics used in protein crystallography.
• To analyze the overall quality of structures in the PDB using nine different metrics.

Main Methods:

• Analysis of all structures in the Protein Data Bank (PDB).
• Evaluation based on nine different quality metrics, including R factor, R(free), and Ramachandran violations.
• Multivariate statistical analysis to identify quality determinants.

Main Results:

• Technological improvements have increased structure determination numbers but not overall structure quality.
• Structures from structural genomics initiatives generally exhibit better quality than those from individual labs.
• A significant association exists between structure quality and publication journal, with high-impact general science journals publishing lower-quality structures.

Conclusions:

• Publication pressure in high-impact journals may contribute to the dissemination of poor-quality protein structures.
• The quality of deposited protein structures has not improved despite technological advancements.
• Structural genomics initiatives appear to produce higher-quality structures compared to individual research laboratories.