Table 1.
Node feature inputs.
Fig 1.
The three types of graphs used in EIGN.
Fig 2.
(A) Overall framework of EIGN. (B) Inter-molecular message passing structure.
Table 2.
Experimental results on the PDBbind core set.
Fig 3.
Scatter plots of predicted values (y-axis) versus actual values (x-axis) for EIGN on the validation set, CASF-2013, and CASF-2016.
Table 3.
Comparison of predicted results between EIGN and state-of-the-art methods on CASF-2013 and CASF-2016.
Fig 4.
Prediction performance of EIGN and four variants on CASF-2013 and CASF-2016.
Fig 5.
Comparison of prediction performance of EIGN, GIGN, and IGN on the CSAR-HIQ-set.
Fig 6.
Contribution analysis of the node feature components in the model.
(A) Ablation study. (B) GNNExplainer.
Fig 7.
The impact of the proportion of similar samples in the training set on model prediction performance.
Fig 8.
Visualization of the performance of EIGN, GIGN, and IGN on the PYGM target.
(A) Experimentally resolved protein structure. (B) Alphafold3-predicted protein structure.
Fig 9.
Scatter plots of the prediction results of EIGN, GIGN, and IGN on the CDK2 target.