Graph Masked Autoencoders (GMAEs) have emerged as a notable self-supervised learning approach for graph-structured data. Existing GMAE models primarily focus on reconstructing node-level information, categorizing them as single-scale GMAEs. This methodology, while effective in certain contexts, tends to overlook the complex hierarchical structures inherent in many real-world graphs. For instance, molecular graphs exhibit a clear hierarchical organization in the form of the atoms-functional groups-molecules structure. Therefore, the inability of single-scale GMAE models to incorporate these hierarchical relationships often results in an inadequate capture of crucial high-level graph information, leading to a noticeable decline in performance. To address this limitation, we propose Hierarchical Graph Masked AutoEncoders (Hi-GMAE), a novel multi-scale GMAE framework designed to handle the hierarchical structures within graphs. First, Hi-GMAE constructs a multi-scale graph hierarchy through graph pooling, enabling the exploration of graph structures across different granularity levels. To ensure masking uniformity of subgraphs across these scales, we propose a novel coarse-to-fine strategy that initiates masking at the coarsest scale and progressively back-projects the mask to finer scales. Furthermore, we integrate a gradual recovery strategy with the masking process to mitigate the learning challenges posed by completely masked subgraphs. Diverging from the standard graph neural network (GNN) used in GMAE models, Hi-GMAE modifies its encoder and decoder into hierarchical structures. This entails using GNN at the finer scales for detailed local graph analysis and employing a graph transformer at coarser scales to capture global information. Such a design enables Hi-GMAE to effectively capture the multi-level information inherent in complex graph structures. Our experiments on 17 graph datasets, covering two graph learning tasks, consistently demonstrate that Hi-GMAE outperforms 29 state-of-the-art self-supervised competitors in capturing comprehensive graph information.

Retrosynthesis prediction aims to infer the precursor compounds and synthetic pathways for the given product molecule. Despite the emergence of numerous data-driven template-free approaches, most rely on single-modality inputs (e.g., SMILES or molecular graphs), limiting them to incorporate multimodal information. Besides, existing methods require complex alignment preprocessing for models to capture unaltered molecular structures in reactions, which rigidly restricts the data format and hinders generalizing to practical application scenarios. To address these challenges, this paper proposes a novel end-to-end template-free learning framework, named Cross-modal Fine-grained Contrastive Learning for Retrosynthesis Prediction (CFC-Retro). By taking both sequence and graph data as inputs, CFC-Retro leverages the cross-attention mechanism to effectively integrate multimodal information during encoding. Additionally, CFC-Retro employs multimodal hybrid-enhanced decoding to produce accurate results. Moreover, the contrastive learning strategy in CFC-Retro implicitly aligns multimodal features at the atomic level. This strategy guides the model to perceive molecular structural changes in chemical reactions, while exhibiting high flexibility. Experiments show that CFC-Retro achieves top-1 accuracy of 53.7% and validity of 99.7% on USPTO-50k with reaction class unknown, underscoring its robust performance in retrosynthesis prediction.