Unverified50% confidenceFactExact time
A study proposes encoding molecules as fragment-level graphs, enabling any message-passing GNN and post-hoc attribution method to provide explanations at the functional-group granularity.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/22/2026
First Seen
Valid until: 1/20/2027
Sources
Related Entities
Related Claims
UnverifiedMolecular generation and optimization can leverage graph neural networks (GNN), variational autoencoders (VAE), or diffusion models to generate entirely new molecular structures with specific properties.77% similarUnverifiedGraph Neural Networks (GNNs) can directly handle non-Euclidean data such as molecular structures, knowledge graphs, and social networks by performing message passing and aggregation on graph structures.77% similarUnverifiedThe upper bound on the expressive power of GNNs is closely related to the graph isomorphism problem, and the Weisfeiler-Leman (WL) graph isomorphism test provides a theoretical framework for characterizing the expressive power of GNNs.70% similarUnverifiedGNoME基于图神经网络（GNN），通过学习Materials Project等密度泛函理论（DFT）数据库中已知晶体结构稳定性规律进行预测68% similarUnverifiedThe de novo protein design pipeline first uses a diffusion model to generate backbones, then uses a GNN to design sequences.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1084240API
curl https://kongchang.com/api/v1/knowledge/claims/1084240MCP
get_claim(id=1084240)