Comprehensive Approach to Simulating Large Scale Conformational Changes in Biological Systems Utilizing a Path...

Area of Science:

• Computational chemistry and structural biology
• Biomolecular simulations and drug discovery

Background:

• Static protein structures from methods like AlphaFold are insufficient for understanding dynamic biological processes.
• Conventional molecular dynamics (MD) simulations are often too slow for drug design projects requiring microsecond to millisecond timescales.
• Existing restraints for conformational sampling have limitations in simulating complex biological motions.

Purpose of the Study:

• To develop a novel three-stage procedure for constructing realistic all-atom path collective variables (PCVs).
• To introduce a new barrier restraint suitable for complex, conformationally driven biological events like allosteric modulation.
• To demonstrate the utility of the new PCV and restraint within a hybrid sampling framework for drug discovery.

Main Methods:

• Developed an all-atom PCV derived from all-atom MD trajectory frames.
• Introduced a new barrier restraint utilizing a scaled reciprocal function for zero-tolerance violation.
• Implemented the PCV and restraint in a hybrid sampling framework combining well-tempered metadynamics and extended-Lagrangian adaptive biasing force (meta-eABF).

Main Results:

• Successfully applied the method to three pharmaceutically relevant examples: ubiquitin sampling in a complex, stabilizing JAK2-V617F, and inducing an activated STING state.
• Demonstrated the effectiveness of the scaled reciprocal barrier restraint in molecular dynamics simulations.
• Provided statistical analysis of meta-eABF free energy estimates for JAK2-V617F and STING examples.

Conclusions:

• The developed PCV and barrier restraint method significantly enhances conformational sampling for complex biomolecules.
• This approach offers a powerful tool for studying conformationally driven biological events in drug discovery.
• The provided code enables reproducibility and application of the method to similar challenges.