Here are some slides I prepared for a guest lecture for the graduate course “Advanced Computational Biology Methods and Research” at McGill University on the topic of network motif mining. In particular I focus on the possibility of recent advances in graph representation learning to break through old bottlenecks of motif size and variability.