Sequence-Dependent Three Interaction Site Model for Single- and Double-Stranded DNA.

Area of Science:

• Computational Biology
• Biophysics
• Molecular Modeling

Background:

• Developing accurate coarse-grained models is crucial for simulating large DNA systems.
• Existing models often struggle to capture the complex interactions governing DNA behavior.

Purpose of the Study:

• To develop a transferable and robust coarse-grained model for both single-stranded DNA (ssDNA) and double-stranded DNA (dsDNA).
• To accurately predict experimentally measurable DNA properties, including structural, mechanical, and thermodynamic behaviors.

Main Methods:

• A three-interaction-site (TIS) coarse-grained model was developed, representing nucleotides with sites for sugar, phosphate, and base.
• Force constants were determined using Boltzmann inversion on Protein Data Bank structures.
• Stacking interaction parameters were refined using a learning procedure to match experimental dimer melting temperatures.

Main Results:

• The TIS model successfully reproduces experimentally measured salt and sequence-dependent sizes of ssDNA and persistence lengths of poly(dA) and poly(dT).
• It accurately predicts force-extension curves, revealing distinct behaviors for poly(dA) (plateau due to stacking) and poly(dT) (entropic).
• The model also accurately predicts dsDNA persistence lengths and DNA hairpin melting temperatures, including sequence-specific trends.

Conclusions:

• The TIS model demonstrates high transferability and robustness for ssDNA and dsDNA simulations without further parameter adjustments.
• It provides a reliable tool for studying DNA conformational dynamics, mechanical properties, and thermodynamics across various conditions.