Molecular complexes at a glance: automated generation of two-dimensional complex diagrams

Area of Science:

• Computational Biology
• Structural Biology
• Biochemistry

Background:

• Automated generation of molecular complex diagrams is crucial for structural biology, biochemistry, and medicinal chemistry.
• Existing methods may lack efficiency or accuracy in representing complex molecular interactions.
• Molecular complex diagrams are widely used in scientific literature and educational materials.

Purpose of the Study:

• To present a novel algorithmic approach for automatically generating structure diagrams of molecular complexes.
• To develop a computational tool that aids in visualizing protein-ligand interactions.
• To provide a key element for computer applications in structural biology and related fields.

Main Methods:

• A combinatorial optimization strategy is employed to solve parts of the layout problem non-heuristically.
• The algorithm generates diagrams including ligands, interacting amino acids, and hydrophilic interactions.
• Molecules are represented as structure diagrams adhering to chemical nomenclature.

Main Results:

• The method was implemented in the software tool PoseView.
• PoseView was tested on 305 protein-ligand complexes from the Protein Data Bank.
• Collision-free layouts were successfully generated for over 75% of the tested complexes.

Conclusions:

• The developed algorithm and PoseView software offer an effective solution for automated generation of molecular complex diagrams.
• The tool addresses challenges in layout generation, providing accurate and collision-free visualizations.
• PoseView is available as a web service, facilitating its use in scientific research and education.