Overview of refinement procedures within REFMAC5: utilizing data from different sources.

Area of Science:

• Crystallography
• Structural Biology
• Computational Biology

Background:

• Macromolecular structure determination relies on refining models against experimental data.
• Ensuring consistency between structural models and chemical knowledge is crucial.

Purpose of the Study:

• To discuss the refinement of macromolecular structures using REFMAC5 within the CCP4 suite.
• To highlight methods for optimizing structural models with experimental data and prior knowledge.

Main Methods:

• Optimization of a posterior conditional probability distribution for model parameters (atomic coordinates, B factors, etc.).
• Utilizing a library of chemical restraints for stereochemical consistency.
• Employing a Bayesian framework with external restraints for low-resolution or low-data scenarios.

Main Results:

• REFMAC5 optimizes model parameters to align with observed structure factor amplitudes/intensities.
• External restraints enhance refinement using homologous structures, secondary structure information, or NMR data.
• The procedure generates improved electron-density maps for model rebuilding.

Conclusions:

• REFMAC5 provides a robust framework for macromolecular structure refinement.
• The integration of prior knowledge and external data improves model accuracy, particularly in challenging cases.
• Generated electron-density maps facilitate iterative model improvement.