Robustness of atomistic Gō models in predicting native-like folding intermediates

Area of Science:

• Computational Biology
• Biophysics
• Protein Dynamics

Background:

• Gō models are widely used for protein folding simulations due to their native-centric construction.
• Understanding how atomistic details of native structures affect Gō model predictions is crucial for their accurate application.
• The robustness of Gō models in predicting folding intermediates requires thorough investigation.

Purpose of the Study:

• To assess the impact of native structure atomistic details on Gō model-predicted protein folding behavior.
• To investigate the reliability of Gō models in predicting the existence and nature of folding intermediate states.
• To compare the folding pathway of a specific mutant (N47G Spc-SH3) using its native structure against in silico generated alternatives.

Main Methods:

• Discrete molecular dynamics simulations employing a Gō potential with a full atomistic protein representation.
• Equilibrium folding simulations to capture protein dynamics.
• Structural clustering and principal component analysis to analyze folding pathways and identify conformational states.

Main Results:

• Gō model predictions for protein folding pathways are sensitive to the specific atomistic details of the input native structure.
• The presence and characteristics of predicted intermediate states can vary significantly based on subtle differences in the native structure.
• Simulations revealed distinct folding behaviors when comparing the N47G Spc-SH3 mutant's native structure with in silico generated alternatives.

Conclusions:

• Atomistic details encoded in native structures play a critical role in dictating protein folding pathways and intermediate states predicted by Gō models.
• The native-centric approach of Gō models necessitates careful consideration of structural input to ensure reliable predictions.
• Further research is needed to refine Gō models to better account for structural variations and improve their predictive power in protein folding studies.