3D structure and stability prediction of DNA with multi-way junctions in ionic solutions

Area of Science:

• Molecular Biology
• Computational Chemistry
• Biophysics

Background:

• Understanding DNA's three-dimensional (3D) structure and stability is crucial for biological function and drug design.
• Predicting DNA folding from sequence remains a significant challenge in computational biology.

Purpose of the Study:

• To develop and validate an improved coarse-grained (CG) model for *ab initio* prediction of DNA folding and stability.
• To assess the model's accuracy in predicting 3D structures and thermal stability of DNA with multi-way junctions.

Main Methods:

• Developed an improved coarse-grained (CG) model incorporating refined electrostatic potential.
• Employed replica-exchange Monte Carlo simulations and weighted histogram analysis.
• Validated predictions against experimental data for DNA thermal stability and structural accuracy.

Main Results:

• The CG model accurately predicts 3D DNA structures with multi-way junctions (mean RMSD ~8.8 Å).
• It outperforms existing fragment-assembly and AI-based prediction methods.
• The model accurately reproduces DNA junction thermal stability across various sequences and lengths (melting temperature deviation < 5 °C) under different ionic conditions.
• Analysis revealed key intermediate states determine overall junction stability.

Conclusions:

• The improved CG model provides a robust framework for predicting complex DNA architectures.
• Offers mechanistic insights into DNA folding, stability, and function.
• Facilitates advancements in structure-based drug design through accurate DNA structure prediction.