Predicting Network Motif Fingerprints with Graph Neural Networks

Pedro C Vieira, Miguel E. P. Silva, Pedro Manuel Pinto Ribeiro

Keywords: motifs, graph representation learning, synthetic data, significance profiles

TL;DR: Studying the ability of MPNNs to predict motif profiles.

Abstract: Graph Neural Networks (GNNs) are a predominant method for graph representation learning. However, beyond subgraph frequency estimation, their application to network motif prediction remains underexplored, with no established benchmarks in the literature. We propose to address this problem, framing motif prediction as an extension of subgraph frequency estimation. Our approach formulates motif estimation as a multitarget regression problem, optimising for interpretability and improving stability and scalability on large graphs. We validate our method using a large synthetic dataset generated by graph generators that mimic real-world data, and further test it on real-world graphs. Our experiments reveal that 1-WL limited models trained on synthetic data struggle to predict accurately motif profiles of real-world networks. However, apart from their reasonable performance within synthetic data, they can generalise to approximate the graph generation processes of real-world networks by comparing their predicted motif profiles with the ones originating from synthetic data. This first study on GNN-based motif prediction sets a benchmark and should open pathways for further developing the connection between motif profiles and subgraph frequency from a graph representation learning perspective.

Primary Area: learning on graphs and other geometries & topologies

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 4513