SE(3)-Equivariant Ternary Complex Prediction Towards Target Protein Degradation

2025-02-26

AI Summary

Accurate prediction of ternary complex structures (E3 ligase–target protein–degrader) in targeted protein degradation (TPD) remains challenging due to mechanistic ambiguity, structural complexity, and severe scarcity of experimentally resolved training data. Method: We propose TernaryFold, the first SE(3)-equivariant graph neural network framework for de novo ternary structure prediction. It integrates intra- and inter-complex ternary attention mechanisms with a query-based pocket point-cloud decoder, enabling end-to-end, prior-free modeling. We also curate TernaryDB, a high-quality, manually validated database of ternary complexes. Contributions/Results: On the PROTAC benchmark, TernaryFold achieves state-of-the-art accuracy and inference speed. In blind docking, it significantly improves molecular glue degradation (MGD) prediction accuracy. Predicted buried surface area strongly correlates with experimental degradation efficacy (r > 0.85), providing an interpretable, structure-guided foundation for rational TPD design.

Technology Category

• Machine Learning: Structured Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Targeted protein degradation (TPD) induced by small molecules has emerged as a rapidly evolving modality in drug discovery, targeting proteins traditionally considered"undruggable". Proteolysis-targeting chimeras (PROTACs) and molecular glue degraders (MGDs) are the primary small molecules that induce TPD. Both types of molecules form a ternary complex linking an E3 ligase with a target protein, a crucial step for drug discovery. While significant advances have been made in binary structure prediction for proteins and small molecules, ternary structure prediction remains challenging due to obscure interaction mechanisms and insufficient training data. Traditional methods relying on manually assigned rules perform poorly and are computationally demanding due to extensive random sampling. In this work, we introduce DeepTernary, a novel deep learning-based approach that directly predicts ternary structures in an end-to-end manner using an encoder-decoder architecture. DeepTernary leverages an SE(3)-equivariant graph neural network (GNN) with both intra-graph and ternary inter-graph attention mechanisms to capture intricate ternary interactions from our collected high-quality training dataset, TernaryDB. The proposed query-based Pocket Points Decoder extracts the 3D structure of the final binding ternary complex from learned ternary embeddings, demonstrating state-of-the-art accuracy and speed in existing PROTAC benchmarks without prior knowledge from known PROTACs. It also achieves notable accuracy on the more challenging MGD benchmark under the blind docking protocol. Remarkably, our experiments reveal that the buried surface area calculated from predicted structures correlates with experimentally obtained degradation potency-related metrics. Consequently, DeepTernary shows potential in effectively assisting and accelerating the development of TPDs for previously undruggable targets.

Problem

• Predicts ternary complex structures
• Targets undruggable protein degradation
• Uses SE(3)-equivariant GNN for accuracy

Innovation

• SE(3)-equivariant GNN for structure prediction
• Encoder-decoder architecture for ternary complexes
• Query-based Pocket Points Decoder