Docking for molecules that bind in a symmetric stack with SymDOCK.

Area of Science:

• Computational chemistry and structural biology
• Neuroscience and drug discovery

Background:

• Identifying ligands for tau protein amyloid fibrils is crucial for understanding neurodegenerative diseases.
• Ligands often bind in symmetrical stacks within tau fibrils, interacting primarily with each other rather than the protein.

Approach:

• Developed SymDOCK, a computational method to dock molecules that exploit the rotational and translational symmetry of tau fibrils.
• SymDOCK generates symmetrical ligand stacks, assesses for clashes, and incorporates ligand-ligand van der Waals energy into docking scores.
• Utilizes ANI, a neural network-based quantum-mechanical method, to evaluate ligand stacking energies.

Key Points:

• SymDOCK successfully reproduced known poses of tau PET tracers in retrospective studies.
• Prospectively predicted the binding structure of the PET tracer MK-6240 to AD PHF tau fibrils before experimental determination.
• Demonstrated enrichment of known ligands over decoys in screens, maintaining high docking speed.

Conclusions:

• SymDOCK is an effective tool for identifying novel symmetrical stacking ligands for amyloid fibrils.
• The method accelerates the discovery of potential therapeutics and diagnostic agents for tauopathies.
• Future applications include large-scale library screening for new symmetrical stackers.