Quantum Mechanical-Cluster Approach to Solve the Bioisosteric Replacement Problem in Drug Design

Area of Science:

• Computational chemistry
• Drug discovery
• Medicinal chemistry

Background:

• Bioisosteric replacements are crucial in drug design for optimizing compound properties.
• Standard computational methods struggle to accurately predict binding affinities for similar bioisosteres due to their shape similarity.

Purpose of the Study:

• To develop an advanced computational methodology for predicting the impact of bioisosteric substitutions on drug activity.
• To specifically address the challenge of H → F replacements in drug design.

Main Methods:

• A quantum mechanical (QM)-cluster approach utilizing the GFN2-xTB semi-empirical method was designed.
• The methodology was applied to evaluate H → F bioisosteric replacements.

Main Results:

• The QM-cluster approach achieved a standard deviation of 0.60 kcal/mol in predicting biological activity changes.
• This performance surpasses the ChemPLP scoring function (0.83 kcal/mol) and approaches experimental accuracy (∼0.42 kcal/mol).

Conclusions:

• The developed QM-cluster method offers accurate and efficient prediction of bioisosteric effects in drug design.
• Its speed and lack of tunable parameters make it a valuable tool for modern drug research.