DEMO-EMol: modeling protein-nucleic acid complex structures from cryo-EM maps by coupling chain assembly with map...

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Accurate atomic structure determination of macromolecular complexes is essential for understanding biological function.
• Cryo-electron microscopy (cryo-EM) is a powerful technique for visualizing these structures, but modeling complexes remains challenging.

Purpose of the Study:

• To introduce DEMO-EMol, an improved computational server for accurate atomic structure modeling of protein-nucleic acid complexes from cryo-EM data.
• To enhance the assembly of complex structures by integrating deep learning-based segmentation and chain fitting.

Main Methods:

• DEMO-EMol utilizes deep learning for segmenting protein and nucleic acid regions within cryo-EM density maps.
• The server assembles complex structures by fitting independently modeled chains into segmented maps, followed by domain-level refinement.
• Quality assessments at overall and residue levels are provided for the assembled models.

Main Results:

• DEMO-EMol demonstrated superior performance in modeling both protein-nucleic acid and protein-protein complexes compared to existing state-of-the-art methods.
• The server's effectiveness was validated across a diverse benchmark set of cryo-EM maps with varying resolutions (1.96–12.77 Å).

Conclusions:

• DEMO-EMol significantly improves the accuracy and efficiency of atomic structure modeling for macromolecular complexes from cryo-EM data.
• The developed web server provides a valuable resource for researchers in structural biology and related fields.