A log-likelihood-gain intensity target for crystallographic phasing that accounts for experimental error

Area of Science:

• Crystallography
• Structural Biology
• Biophysics

Background:

• Crystallographic experiments measure Bragg intensities, but phasing calculations require structure-factor amplitudes.
• Converting intensities to amplitudes is complicated by experimental errors, especially negative intensities, leading to significant defects in current methods.

Purpose of the Study:

• To develop a novel method for handling experimental errors in crystallographic data.
• To improve the accuracy and reliability of macromolecular crystallography phasing and refinement.

Main Methods:

• Formulation of a log-likelihood-gain function (LLGI) that directly uses Bragg intensities and their experimental error estimates.
• LLGI is designed to handle large experimental errors by appropriately downweighting problematic reflections without introducing bias.

Main Results:

• LLGI demonstrates correct asymptotic behavior for data with large experimental errors.
• This approach eliminates the need for the French and Wilson method for converting intensities to amplitudes in likelihood calculations.

Conclusions:

• LLGI offers a robust solution for incorporating experimental error into crystallographic phasing and refinement.
• It is broadly applicable to various macromolecular crystallography algorithms and suggests depositing intensity data in the Protein Data Bank.