Comprehensive Evaluation and Optimization of Martini Simulations for Dipeptide Self-Assembly.

Area of Science:

• Computational chemistry
• Biophysics
• Materials science

Background:

• Predicting peptide self-assembly using coarse-grained molecular dynamics (CGMD) simulations is difficult due to parameter sensitivity.
• The Martini force field is widely used for CGMD but requires careful parameterization for accurate peptide self-assembly prediction.

Purpose of the Study:

• To develop and validate a computational-experimental framework for optimizing Martini-based CGMD simulations of dipeptide self-assembly.
• To identify key simulation parameters influencing the accuracy of predicting peptide aggregation propensity (AP).

Main Methods:

• Systematic analysis of 40 chemically diverse dipeptides using CGMD simulations.
• Evaluation of parameters including simulation time, peptide concentration, system size, backbone bead type assignment (H/E/C), and terminal charges.
• Experimental validation using transmission electron microscopy (TEM).

Main Results:

• Secondary structure and terminal charge significantly impact dipeptide aggregation behavior by altering coarse-grained particle types and interactions.
• The influence of secondary structure and terminal charge decreases with increasing peptide length.
• Simulations using uncharged termini and beta-sheet secondary structure assignment demonstrated the highest predictive accuracy, confirmed by TEM.

Conclusions:

• Optimized Martini simulation parameters, specifically uncharged termini and beta-sheet assignment, enhance the reliability of predicting short-peptide self-assembly.
• This framework provides mechanistic insights into parameter-dependent variability in CGMD simulations.
• The study identified novel self-assembling dipeptide candidates through comprehensive screening with optimized parameters.