Structure centric graph model speeds and improves cross domain learning
SCGFM-ART: Amortized Relational Transport for Structure-Centric Graph Foundation Models
Machine LearningArtificial Intelligence
Summary
Graphs are structures that represent connections between things, but graphs from different areas can be very different, making it hard to analyze them together. The authors created SCGFM-ART, a method that aligns different graphs into a shared framework by relating their parts to a common set of landmarks. This approach makes understanding and transferring information from one type of graph to another faster and more accurate. Their method also avoids slow calculations during use, enabling much quicker results on new graphs.
What this means in practice
- •For machine learning engineers: Enable faster and more accurate cross-domain graph data analysis by directly predicting alignment plans without costly runtime computations.
- •For bioinformatics teams: Improve the transfer of structural patterns from known biological networks to less studied ones, enhancing disease or drug interaction models.
Authors
Xiaodong He, Xincheng Wang, Zhao Kang
Abstract
Graph foundation models (GFMs) aim to learn transferable representations across severely heterogeneous graph domains. However, severe domain shifts in topology, graph scale, and feature semantics impede the construction of a unified, domain-agnostic representation space. To address this, we propose SCGFM-ART, a structure-centric GFM framework that aligns arbitrary graphs onto a shared relational atlas via Amortized Relational Transport (ART). The relational atlas serves as a universal coordinate system defined by a finite set of relational landmarks (bases), while ART directly predicts reusable, end-to-end graph-to-base transport plans, bypassing costly runtime Gromov-Wasserstein optimizations. Under this formulation, SCGFM-ART decomposes a graph into a unified representation: globally via its relational response coordinates relative to the atlas, and locally via its node-to-role structural correspondences. These correspondences project disparate node attributes into a canonical role space, resolving structural and semantic heterogeneity within a singular alignment interface. Rigorously modeling graphs and atlas bases as finite measured relational spaces, we establish coordinate fidelity bounds, prove stability under predicted transport plans, and derive an amortized coverage bound that guarantees our learning objective tightly surrogates ideal relational coverage. Benchmarked across 14 cross-domain graph- and node-level classification tasks, SCGFM-ART achieves state-of-the-art transferability, securing superior average ranks of 2.29 and 1.14, respectively. Topological perturbation analyses demonstrate that node-role transport retains fine-grained structural nuances beyond global coordinates. On real-world benchmarks, the amortized formulation yields 44.2 to 85.1 times faster frozen target-domain inference by avoiding iterative alignment at test time.