Computer simulation of antibody binding specificity

Area of Science:

• Computational Biology
• Structural Biology
• Immunology

Background:

• Antibody-protein interactions are crucial in immunology and drug development.
• Accurate prediction of protein-antibody docking remains a challenge.
• Existing methods often focus on maximizing surface contact, which may not reflect binding energy accurately.

Purpose of the Study:

• To develop a novel Monte Carlo algorithm for predicting optimal docking configurations of hen egg white lysozyme to antibodies.
• To investigate the effectiveness of energy minimization using coarse-grained potentials for protein-antibody docking.
• To compare the performance of different dielectric functions in docking simulations.

Main Methods:

• A rigid-body Monte Carlo algorithm was employed to search for docking configurations.
• Coarse-grained pair potentials were used to compute interaction energies.
• All-atom OPLS potentials were utilized for final refinement of top solutions.
• An exponential distance-dependent dielectric function was tested against a linear form.

Main Results:

• The algorithm successfully identified the native conformation as the preferred solution for three different antibodies.
• Minimizing energy with coarse-grained potentials proved effective for docking.
• The use of an exponential distance-dependent dielectric function improved simulation accuracy.
• The method avoids explicit maximization of surface contact.

Conclusions:

• The developed Monte Carlo algorithm provides an accurate method for predicting protein-antibody docking.
• Energy minimization using coarse-grained potentials is a viable strategy for computational docking.
• The choice of dielectric function significantly impacts the accuracy of docking simulations.