Overview of GraSSRep. (A) Reads are assembled into contigs, forming the nodes of the assembly graph. The graph structure (edges) is provided by metaSPAdes. Also, feature vectors are computed for each contig. (B) Contigs with distinctive sequencing features are selected as training nodes and labeled. (C) The assembly graph is input into a GNN. Embeddings are generated for each contig and combined with the initial features. A random forest classifier predicts labels for all contigs based on the augmented feature vectors. (D) Sequencing features are employed to identify outliers within each predicted class, leading to the reassignment of their class labels.