Modelling prior distributions of atoms for macromolecular refinement and completion

Area of Science:

• Crystallography
• Structural Biology
• Computational Chemistry

Background:

• Macromolecular structures are often refined with incomplete atomic models.
• Modeling missing atoms and bulk solvent is crucial for accurate structure determination.
• Current methods may struggle with low-resolution data and uncertain features.

Purpose of the Study:

• To describe techniques for modeling missing atoms and bulk solvent in macromolecular crystallography.
• To introduce positional probability distributions for structure refinement.
• To propose homographic exponential modeling for improved envelope approximation and ab initio phasing.

Main Methods:

• Utilizing positional probability distributions to model missing atomic coordinates.
• Employing electron-density maps or tentative models as starting information.
• Developing and applying homographic exponential modeling for macromolecular envelopes.

Main Results:

• Positional probability distributions allow retention of low-resolution phase information.
• Homographic exponential modeling provides a superior approximation of protein envelopes compared to Fourier expansion.
• The new methods facilitate structure completion and refinement with incomplete data.

Conclusions:

• The described techniques enhance the modeling of missing structural components and bulk solvent.
• Homographic exponential modeling offers a robust and efficient alternative for envelope representation.
• These advancements hold potential for improving ab initio phasing strategies in structural biology.