Accurate Physics-Based Flexible Docking of Macrocyclic Ligands

Area of Science:

• Computational chemistry
• Medicinal chemistry
• Drug discovery

Background:

• Macrocycles are a vital therapeutic class facing computational modeling challenges.
• Existing methods struggle with sampling bioactive conformations and predicting binding modes.
• These limitations hinder structure-based drug discovery for macrocyclic therapeutics.

Purpose of the Study:

• To develop an improved computational docking protocol for macrocycles.
• To enhance the accuracy of predicting macrocycle binding modes and conformations.
• To overcome limitations in current macrocycle modeling techniques.

Main Methods:

• Developed a novel flexible macrocycle docking protocol.
• Integrated existing Schrödinger macrocycle sampling and small molecule docking technologies.
• Benchmarked the protocol using an expanded dataset of 240 receptor-macrocycle systems.

Main Results:

• The new protocol significantly outperforms existing docking tools.
• Achieved 82% accuracy in recapitulating receptor-bound structures within the top 2 poses.
• Successfully modeled binding for several clinically relevant macrocyclic compounds.

Conclusions:

• The developed flexible macrocycle docking protocol shows significant promise for structure-based drug discovery.
• The method improves the prediction of macrocycle binding poses and conformations.
• Further research is needed to address challenges like induced-fit effects and strain in macrocycle modeling.