aiqbench analysis
CORAL introduces a reinforcement learning framework specifically designed to solve the unique docking challenges posed by amyloid fibrils, which differ significantly from standard globular proteins. By incorporating cooperative ligand-ligand stacking energy into its reward function, this model addresses data scarcity and geometric complexity in neurodegenerative disease research without relying on extensive supervised datasets.
CORAL: Learning Amyloid Fibril Ligand Docking with Cooperative Binding Rewards
arXiv:2607.17412v1 Announce Type: new Abstract: A hallmark of neurodegenerative diseases such as Alzheimer's and Parkinson's is the aberrant aggregation of proteins into amyloid fibrils, and small molecules that selectively bind to these fibrils hold promise as diagnostics, imaging probes, and therapeutics. Predicting how such ligands bind to fibril targets, however, presents two fundamental challenges. First, resolved co-crystal structures of amyloid-ligand complexes are exceptionally scarce; even with recent advances in cryo-EM only a handful have been structurally characterized, making supervised training of docking models impractical for this target class. Second, amyloid fibrils present a binding mode fundamentally different from globular proteins: ligands intercalate into longitudinal cross-$\beta$ grooves and stack cooperatively along the fibril axis, a geometry that existing docking models are not designed to capture. To address these challenges, we present CORAL (COopeRative Amyloid Ligand docking), a reinforcement learning framework that trains a generative docking model to produce ligand pose distributions tailored to the cross-$\beta$ groove geometry. Our reward explicitly incorporates cooperative ligand-ligand stacking energy alongside protein-ligand docking affinity, directly capturing the distinctive binding geometry of amyloid fibrils. We further introduce a curated evaluation set of amyloid-ligand complexes constructed from model-generated poses validated by domain experts. Experiments on both experimentally resolved structures and this evaluation set demonstrate improved pose quality and binding affinity correlation over existing docking baselines.
Abstract:A hallmark of neurodegenerative diseases such as Alzheimer's and Parkinson's is the aberrant aggregation of proteins into amyloid fibrils, and small molecules that selectively bind to these fibrils hold promise as diagnostics, imaging probes, and therapeutics. Predicting how such ligands bind to fibril targets, however, presents two fundamental challenges. First, resolved co-crystal structures of amyloid-ligand complexes are exceptionally scarce; even with recent advances in cryo-EM only a handful have been structurally characterized, making supervised training of docking models impractical for this target class. Second, amyloid fibrils present a binding mode fundamentally different from globular proteins: ligands intercalate into longitudinal cross-$\beta$ grooves and stack cooperatively along the fibril axis, a geometry that existing docking models are not designed to capture. To address these challenges, we present CORAL (COopeRative Amyloid Ligand docking), a reinforcement learning framework that trains a generative docking model to produce ligand pose distributions tailored to the cross-$\beta$ groove geometry. Our reward explicitly incorporates cooperative ligand-ligand stacking energy alongside protein-ligand docking affinity, directly capturing the distinctive binding geometry of amyloid fibrils. We further introduce a curated evaluation set of amyloid-ligand complexes constructed from model-generated poses validated by domain experts. Experiments on both experimentally resolved structures and this evaluation set demonstrate improved pose quality and binding affinity correlation over existing docking baselines.
Machine Learning (cs.LG)
arXiv:2607.17412 [cs.LG]
(or arXiv:2607.17412v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2607.17412 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 306 words
Related coverage
Harmonic Torsional Diffusion for Protein-Ligand Flexible Docking
arXiv · Aug 24, 2026
GPU-Accelerated Astrodynamics World Models for Spacecraft Rendezvous and Proximity Operations
arXiv · Sep 4, 2026
FuseDiff: Symmetry-Preserving Joint Diffusion for Dual-Target Structure-Based Drug Design
arXiv · Sep 9, 2026
Trustworthy Protein-Ligand Binding Affinity Prediction via Reliability-Aware Multi-Engine Fusion
arXiv · Jul 21, 2026