Enhancing backbone sampling in Monte Carlo simulations using internal coordinates normal mode analysis.

Area of Science:

• Computational Biology
• Biophysics
• Structural Biology

Background:

• Normal mode analysis (NMA) is increasingly used for sampling protein conformational landscapes.
• Existing NMA methods often utilize Cartesian coordinates, which may not fully capture internal protein dynamics.

Purpose of the Study:

• To implement and evaluate an internal coordinate normal mode analysis (IC-NMA) method.
• To explore protein flexibility using IC-NMA coupled with Monte Carlo simulations (PELE).
• To compare the efficiency and accuracy of IC-NMA against Cartesian coordinate NMA and molecular dynamics (MD) simulations.

Main Methods:

• Developed an internal coordinate normal mode analysis method.
• Integrated IC-NMA with the Monte Carlo simulation package PELE.
• The method involves alternating backbone perturbation via torsional normal modes and side-chain resampling.
• Validated the approach using ubiquitin and c-Src kinase as test systems.
• Compared results against anisotropic network model (ANM) and reference molecular dynamics simulations.

Main Results:

• The internal coordinate approach sampled phase space closer to molecular dynamics simulations than the Cartesian coordinate ANM.
• The new IC-NMA method demonstrated a significant speedup of approximately 5-7×.
• The method effectively explores protein conformational flexibility.

Conclusions:

• Internal coordinate normal mode analysis provides a more accurate representation of protein conformational dynamics compared to Cartesian methods.
• The developed IC-NMA method offers a computationally efficient alternative for exploring protein flexibility.
• This approach is a promising candidate for future NMA implementations in Monte Carlo simulations.