AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

FuseMine: Robust Multi-Modal Compound-Protein Interaction Prediction via Differential Attention Feature Mining

Junlin Xu, Zhuang Zhang, Zhenghang Gong, Jincan Li, Pan Zeng, Zilong Zhang, Xiong Li, Shuting Jin, Haowen Chen, Yajie Meng

AAAI 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 1 citations (OpenAlex)

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 changedWhereMeasuredEffect
remove MFOA multi-modal fusion module (ablation)FuseMine · BindingDB and BioSNAP (cold-pair)AUC / AUPRImproved
remove RDFM differential attention module (ablation)FuseMine · BindingDB and BioSNAP (cold-pair)AUC / AUPRImproved
FuseMine vs MlanDTI on small datasetsFuseMine · Human (random split)AUC: 0.988 -> 0.989No clear effect
FuseMine vs MlanDTI on larger datasetFuseMine · BindingDB (random split)AUC: 0.945 -> 0.968Improved
FuseMine vs MlanDTI under cold-pair splitFuseMine · BindingDB (cold-pair)AUC: 0.671 -> 0.745Improved
FuseMine vs MlanDTI under cold-pair splitFuseMine · BioSNAP (cold-pair)F1: 0.653 -> 0.611Hurt
FuseMine vs MlanDTI (pseudo-labeling) cross-domainFuseMine · BindingDB (cross-domain)AUC: 0.687 -> 0.657No clear effect
add pseudo-labeling to MlanDTIMlanDTI · BindingDB (cross-domain)AUC: 0.657 -> 0.687Improved
add adversarial domain adaptation (CDAN) to DrugBANDrugBAN · BioSNAP (cross-domain)AUC: 0.608 -> 0.673Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy