De novo Generation for Molecular Structure Elucidation from Mass Spectrometry
Microsoft researchers propose Diffs, a novel generative model using diffusion models and graph neural networks to infer molecular structures from mass spectrometry data, significantly improving accuracy and efficiency.
入选理由:Diffs 模型结合了扩散模型和图神经网络,能够从质谱数据中生成分子结构,准确率达到 90% 以上。

