NEAT-DNA: A Chemically Accurate, Sequence-Dependent Coarse-Grained Model for Large-Scale DNA Simulations.

Area of Science:

• Computational Biology
• Structural Genomics
• Biophysics

Background:

• Simulating DNA behavior at biologically relevant scales is computationally challenging.
• Existing coarse-grained DNA models often lack chemical accuracy or produce unphysical conformations.
• Developing efficient and physically realistic DNA models is crucial for studying genome organization.

Purpose of the Study:

• Introduce NEAT-DNA, a novel coarse-grained DNA model.
• Address limitations in physical realism and parameter optimization of previous models.
• Enable accurate and efficient large-scale DNA simulations.

Main Methods:

• Developed a physically principled energy formulation for DNA.
• Integrated atomistic simulations and experimental data into a unified training framework.
• Created NEAT-DNA, a computationally efficient and sequence-specific coarse-grained model.

Main Results:

• NEAT-DNA accurately reproduces sequence-dependent DNA structure and flexibility.
• The model demonstrates high fidelity while maintaining computational efficiency.
• Achieved significant advances over previous models lacking sequence specificity or introducing distortions.

Conclusions:

• NEAT-DNA offers a high-fidelity, tractable DNA representation for chromatin folding studies.
• Provides a foundation for large-scale simulations linking molecular detail to chromatin organization.
• Opens new avenues for predictive modeling in structural genomics and computational biology.