Computational Approaches for Protein-DNA/RNA Complex Modeling for Cryo-EM Maps

Area of Science:

• Structural Biology
• Computational Biology
• Biophysics

Background:

• Cryo-electron microscopy (cryo-EM) is pivotal for determining macromolecular structures.
• Existing computational tools primarily focus on protein structures, with limited options for protein-nucleic acid complexes.
• Accurate atomic modeling of these complexes is crucial for understanding biological mechanisms.

Purpose of the Study:

• To introduce and demonstrate the usage of ComplexModeler and CryoZeta for modeling protein-nucleic acid complexes using cryo-EM data.
• To provide accessible computational tools through the EMSuite web server for the structural biology community.
• To facilitate the generation of high-resolution atomic models for complex biological assemblies.

Main Methods:

• ComplexModeler integrates DiffModeler (diffusion model for backbone tracing) and CryoREAD (nucleotide identification and atomic model construction).
• CryoZeta employs a diffusion-based generative model combining sequence-based prediction with cryo-EM density features.
• Both tools are utilized via the user-friendly EMSuite web server, offering structure modeling and validation.

Main Results:

• Demonstrated successful application of ComplexModeler and CryoZeta for modeling protein-DNA/RNA complexes.
• Achieved accurate atomic model construction from cryo-EM density maps, even at resolutions up to 5 Å.
• Validated the efficacy of diffusion models and integrated approaches for complex structural determination.

Conclusions:

• ComplexModeler and CryoZeta offer robust solutions for modeling protein-nucleic acid complexes from cryo-EM data.
• The EMSuite web server provides a centralized and accessible platform for these advanced modeling tools.
• These advancements enhance the capabilities of structural biologists in characterizing intricate molecular machinery.