Timezone: »

Subgraph Neural Networks
Emily Alsentzer · Samuel Finlayson · Michelle Li · Marinka Zitnik

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1432

Deep learning methods for graphs achieve remarkable performance on many node-level and graph-level prediction tasks. However, despite the proliferation of the methods and their success, prevailing Graph Neural Networks (GNNs) neglect subgraphs, rendering subgraph prediction tasks challenging to tackle in many impactful applications. Further, subgraph prediction tasks present several unique challenges: subgraphs can have non-trivial internal topology, but also carry a notion of position and external connectivity information relative to the underlying graph in which they exist. Here, we introduce SUBGNN, a subgraph neural network to learn disentangled subgraph representations. We propose a novel subgraph routing mechanism that propagates neural messages between the subgraph’s components and randomly sampled anchor patches from the underlying graph, yielding highly accurate subgraph representations. SUBGNN specifies three channels, each designed to capture a distinct aspect of subgraph topology, and we provide empirical evidence that the channels encode their intended properties. We design a series of new synthetic and real-world subgraph datasets. Empirical results for subgraph classification on eight datasets show that SUBGNN achieves considerable performance gains, outperforming strong baseline methods, including node-level and graph-level methods, by 19.8% over the strongest baseline. SUBGNN performs exceptionally well on challenging biomedical datasets where subgraphs have complex topology and even comprise multiple disconnected components.

Author Information

Emily Alsentzer (MIT)
Samuel Finlayson (Harvard Medical School)

Samuel Finlayson is a MD-PhD Candidate studying jointly at Harvard Medical School and Massachusetts Institute of Technology. His research focuses on developing machine learning methods for precision medicine. Current applications focus on neurological diseases and extend techniques from computer vision, natural language processing, and single-cell genomics. Previously, he studied Biomedical Informatics at Stanford University.

Michelle Li (Harvard Medical School)
Marinka Zitnik (Harvard University)

More from the Same Authors