RISE: Radius of Influence based Subgraph Extraction for 3D Molecular Graph Explanation

2025-05-04

AI Summary

Existing 3D geometric graph neural networks (GNNs) for molecular modeling suffer from limited interpretability, as conventional 2D subgraph-based explanation methods fail to capture spatially grounded atomic interactions. Method: We propose the first 3D structure-aware interpretability framework—Physical-Aware Influence Radius (PAIR)—which defines a geometry- and message-passing-consistent spatial influence region for each atom. PAIR extracts localized 3D subgraphs via distance-weighted importance scoring and differentiable subgraph masking optimization on truncated 3D graphs. Contribution/Results: Evaluated on multiple molecular property prediction benchmarks, PAIR significantly improves explanation fidelity and chemical plausibility over prior methods. Visualizations align with domain-expert knowledge, demonstrating that PAIR overcomes the fundamental limitation of 2D explanation paradigms in 3D molecular learning.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

3D Geometric Graph Neural Networks (GNNs) have emerged as transformative tools for modeling molecular data. Despite their predictive power, these models often suffer from limited interpretability, raising concerns for scientific applications that require reliable and transparent insights. While existing methods have primarily focused on explaining molecular substructures in 2D GNNs, the transition to 3D GNNs introduces unique challenges, such as handling the implicit dense edge structures created by a cut-off radius. To tackle this, we introduce a novel explanation method specifically designed for 3D GNNs, which localizes the explanation to the immediate neighborhood of each node within the 3D space. Each node is assigned an radius of influence, defining the localized region within which message passing captures spatial and structural interactions crucial for the model's predictions. This method leverages the spatial and geometric characteristics inherent in 3D graphs. By constraining the subgraph to a localized radius of influence, the approach not only enhances interpretability but also aligns with the physical and structural dependencies typical of 3D graph applications, such as molecular learning.

Problem

• Explaining 3D GNNs' limited interpretability in molecular modeling
• Addressing dense edge structures from cut-off radius in 3D GNNs
• Localizing node influence radius for spatial-structural interaction insights

Innovation

• Localizes explanation to node's 3D neighborhood
• Assigns radius of influence for message passing
• Leverages 3D spatial and geometric characteristics