Research map / Probabilistic and scientific modeling / Molecules and proteins
S²Drug: Bridging Protein Sequence and 3D Structure in Contrastive Representation Learning for Virtual Screening
Bowei He, Bowen Gao, Yankai Chen, Yanyan Lan, Chen Ma, Philip S. Yu, Ya-Qin Zhang, Wei-Ying Ma
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 bilateral data sampling to ChemBL pretraining (ablation) | S2Drug · DUD-E | AUROC: 88.73 -> 92.46 | Improved |
| add sequence-structure fusion module (ablation) | S2Drug · DUD-E | AUROC: 87.92 -> 92.46 | Improved |
| add auxiliary binding site prediction task (ablation) | S2Drug · DUD-E | AUROC: 89.58 -> 92.46 | Improved |
| add sequence-structure fusion module (ablation) | S2Drug · LIT-PCBA | BEDROC: 6.88 -> 8.69 | Improved |
| add bilateral data sampling (ablation) | S2Drug · LIT-PCBA | AUROC: 56.12 -> 58.23 | Improved |
| add auxiliary binding site prediction (ablation) | S2Drug · LIT-PCBA | AUROC: 56.47 -> 58.23 | Improved |
| set balancing coefficient lambda above 1.0 (ablation) | S2Drug · DUD-E | AUROC | Hurt |
| use extreme temperature values (too low or too high) (ablation) | S2Drug · DUD-E | AUROC | Hurt |
| add protein sequence pretraining and fusion over structure-only contrastive learning (S2Drug vs DrugHash) | contrastive VS · DUD-E | AUROC: 83.73 -> 92.46 | Improved |
| S2Drug vs best baseline on LIT-PCBA AUROC | S2Drug · LIT-PCBA | AUROC: 62.78 -> 58.23 | Hurt |
| train with homology exclusion identity cutoffs (90% to HMM) (ablation) | S2Drug · DUD-E homology exclusion | AUROC | No clear effect |
| joint binding site prediction improves binding site localization vs baselines | S2Drug · HOLO4K | F1 Score: 52.91 -> 51.66 | Hurt |
Closest papers by meaning
- S-MolSearch: 3D Semi-supervised Contrastive Learning for Bioactive Molecule Search (NeurIPS 2024 · similarity 0.89)
- AANet: Virtual Screening under Structural Uncertainty via Alignment and Aggregation (NeurIPS 2025 · similarity 0.88)
- DrugHash: Hashing Based Contrastive Learning for Virtual Screening (AAAI 2025 · similarity 0.87)
- M²N: A Progressive Macro-to-Micro 3D Modeling Scheme for Unveiling Drug-Target Affinity (AAAI 2025 · similarity 0.87)
- 3DMolFormer: A Dual-channel Framework for Structure-based Drug Discovery (ICLR 2025 · similarity 0.87)
- TurboHopp: Accelerated Molecule Scaffold Hopping with Consistency Models (NeurIPS 2024 · similarity 0.87)
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.
