English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Masked graph modeling excels in the self-supervised representation learningof molecular graphs. Scrutinizing previous studies, we can reveal a commonscheme consisting of three key components: (1) graph tokenizer, which breaks amolecular graph into smaller fragments (i.e., subgraphs) and converts them intotokens; (2) graph masking, which corrupts the graph with masks; (3) graphautoencoder, which first applies an encoder on the masked graph to generate therepresentations, and then employs a decoder on the representations to recoverthe tokens of the original graph. However, the previous MGM studies focusextensively on graph masking and encoder, while there is limited understandingof tokenizer and decoder. To bridge the gap, we first summarize popularmolecule tokenizers at the granularity of node, edge, motif, and Graph NeuralNetworks (GNNs), and then examine their roles as the MGM's reconstructiontargets. Further, we explore the potential of adopting an expressive decoder inMGM. Our results show that a subgraph-level tokenizer and a sufficientlyexpressive decoder with remask decoding have a large impact on the encoder'srepresentation learning. Finally, we propose a novel MGM method SimSGT,featuring a Simple GNN-based Tokenizer (SGT) and an effective decodingstrategy. We empirically validate that our method outperforms the existingmolecule self-supervised learning methods. Our codes and checkpoints areavailable at https://github.com/syr-cn/SimSGT.

Rethinking Tokenizer and Decoder in Masked Graph Modeling for Molecules