图神经网络模型预测和解释离子液体毒性的研究
冯海军, 章冰璇, 周健

Predicting and interpreting the toxicity of ionic liquids using graph neural network
Haijun FENG, Bingxuan ZHANG, Jian ZHOU
图9 4个数据集中正负毒性贡献最高的前15个原子基团的权重分布
Fig.9 Distributions of top 15 atomic weights for both positive and negative contributions in 4 datasets