On the variability of experimental data in macromolecular crystallography

Area of Science:

• Structural biology
• Crystallography

Background:

• Data processing in macromolecular crystallography often underestimates experimental errors.
• Discrepancies exist between expected and observed R values in crystallographic studies.

Purpose of the Study:

• To compare experimental error estimates from data-processing algorithms with those from multiple crystal data collection.
• To identify sources of error inflation in macromolecular crystallography.

Main Methods:

• Utilized a multiple crystal data-collection protocol.
• Compared direct error estimates with algorithm-determined errors.
• Analyzed data from multiple sections of the same crystal.

Main Results:

• Several-fold error inflation is required to account for crystal-to-crystal variation.
• Non-uniform crystal growth contributes significantly to data variation.
• Underestimation of errors is more pronounced at lower resolutions and higher signal-to-noise ratios.

Conclusions:

• Crystal-to-crystal and within-crystal variations are major sources of underestimated experimental errors.
• Differential X-ray absorption and unit-cell parameter variations also contribute to error inflation.
• Addressing these errors is crucial for improving the accuracy of macromolecular crystallography models.