Representing stereochemical information in macromolecular electron-density distributions by multi-dimensional...

Area of Science:

• Crystallography
• Structural Biology
• Computational Chemistry

Background:

• Conventional electron-density histograms are valuable targets but treat density values independently.
• Existing methods do not account for relationships between neighboring electron density values.

Purpose of the Study:

• To introduce and evaluate a multi-dimensional histogram incorporating electron density derivatives.
• To assess the utility of this enhanced histogram for improving crystallographic phase determination and evaluation.

Main Methods:

• Developed a multi-dimensional histogram using joint frequencies of density values, gradient magnitude, and Laplacian.
• Compared the information content and sensitivity to phase errors of the multi-dimensional histogram against conventional histograms.

Main Results:

• The multi-dimensional histogram demonstrates minimal dependence on molecular folding and packing.
• This method captures significantly more stereochemical information compared to conventional electron-density histograms.
• The gradient histogram component shows heightened sensitivity to crystallographic phase errors.

Conclusions:

• Multi-dimensional histograms offer a more robust representation of electron density, capturing richer stereochemical detail.
• These enhanced histograms can serve as improved targets for density modification in crystallographic refinement.
• The proposed method provides more reliable figures of merit for assessing the correctness of crystallographic phases.