Research map / Probabilistic and scientific modeling / Molecules and proteins
Protein Interface Prediction using Graph Convolutional Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add neighbor convolution (Node Average, 3 layers) versus no convolution (ablation) | residue pair classifier · DBD protein interface test set | AUC: 0.808 -> 0.891 | Helped |
| use Node and Edge Average (2 layers) over PAIRpred SVM | interface prediction · DBD protein interface test set | AUC: 0.863 -> 0.898 | Helped |
| stack 2 rather than 1 convolution layer in Node Average (ablation) | Node Average · DBD protein interface test set | AUC: 0.864 -> 0.882 | Helped |
| stack 3 rather than 2 layers in Node Average (ablation) | Node Average · DBD protein interface test set | AUC: 0.882 -> 0.891 | Helped |
| stack 4 rather than 3 layers in Node Average (ablation) | Node Average · DBD protein interface test set | AUC: 0.891 -> 0.889 | No clear effect |
| add edge features (Eq. 2 vs Eq. 1), 1 layer (ablation) | graph convolution · DBD protein interface test set | AUC: 0.864 -> 0.876 | Helped |
| add edge features (Eq. 2 vs Eq. 1), 3 layers (ablation) | graph convolution · DBD protein interface test set | AUC: 0.891 -> 0.895 | No clear effect |
| use separate weights for center node and neighbors (Node Average vs MFN), 3 layers (ablation) | graph convolution · DBD protein interface test set | AUC: 0.873 -> 0.891 | Helped |
| order-independent Node Average versus order-dependent convolution, 3 layers (ablation) | graph convolution · DBD protein interface test set | AUC: 0.891 -> 0.891 | No clear effect |
| order-dependent convolution with 1 layer versus 3 layers (ablation) | Order Dependent · DBD protein interface test set | AUC: 0.854 -> 0.891 | Helped |
| DTNN versus Node and Edge Average at 2 layers (ablation) | DTNN · DBD protein interface test set | AUC: 0.898 -> 0.880 | Hurt |
| diffusion convolution with 5 hops instead of 2 hops (ablation) | DCNN · DBD protein interface test set | AUC: 0.790 -> 0.828 | Helped |
Closest papers by meaning
- Towards Multiscale Graph-based Protein Learning with Geometric Secondary Structural Motifs (NeurIPS 2025 · similarity 0.86)
- Enhancing Graph Contrastive Learning for Protein Graphs from Perspective of Invariance (ICML 2025 · similarity 0.85)
- Evaluating Representation Learning on the Protein Structure Universe (ICLR 2024 · similarity 0.85)
- On the Scalability of GNNs for Molecular Graphs (NeurIPS 2024 · similarity 0.85)
- EquiPocket: an E(3)-Equivariant Geometric Graph Neural Network for Ligand Binding Site Prediction (ICML 2024 · similarity 0.84)
- Clustering for Protein Representation Learning (CVPR 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.