Predicting the Dynamic Viscosity of High-Concentration Antibody Solutions with a Chemically Specific Coarse-Grained...

Area of Science:

• Biophysics
• Computational Chemistry
• Materials Science

Background:

• Protein solution viscosity is crucial for biopharmaceutical formulation.
• Current experimental methods are material-intensive and labor-intensive.
• Atomistic molecular dynamics (MD) simulations are computationally expensive for large systems.

Purpose of the Study:

• To assess the Martini 3 coarse-grained MD (CG-MD) force field's ability to predict viscosity in high-concentration antibody solutions.
• To validate CG-MD as a computationally efficient alternative to atomistic simulations.
• To refine force fields for accurate biopharmaceutical property prediction.

Main Methods:

• Utilized coarse-grained molecular dynamics (CG-MD) simulations with the Martini 3 force field.
• Optimized protein-protein interactions within the Martini 3 framework.
• Incorporated a previously developed Martini 3-exc model for arginine excipients.

Main Results:

• The refined Martini 3 force field accurately predicted elevated viscosities in concentrated F(ab')2 antibody fragment solutions.
• The Martini 3-exc model successfully captured the viscosity-lowering effect of arginine excipients.
• CG-MD simulations demonstrated significant computational savings compared to atomistic MD.

Conclusions:

• Martini 3 CG-MD simulations, with optimized interactions, can reliably predict the viscosity of dense biopharmaceutical solutions.
• This approach offers a computationally feasible method for predicting properties of antibody formulations.
• Physics-based computational predictions can accelerate biopharmaceutical formulation development.