Research map / Vision-language and multimodal / Cross-modal learning
Deciphering Genotype-Phenotype Mechanisms from High-Content Profiling via Knowledge-Guided Multi-modal Graph Learning
Hanjing Lin, Jiahua Rao, Youhan Sun, Jiancong Xie, Yuedong Yang
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 phenotype-driven morphological edges (r_morph) to the knowledge graph (ablation) | KERNEL · hESC GRN inference | AUC / AUPR | Improved |
| add phenotypic interaction contrastive loss on morphological embeddings (ablation) | KERNEL · hESC GRN inference | AUPR, Hit@500 | Improved |
| add neighborhood-overlap structure-aware loss L_sim (ablation) | KERNEL · hESC GRN inference | AUPR, Hit@500 | Improved |
| increase top-K% morphological edges or threshold sigma beyond intermediate values (ablation) | KERNEL · hESC GRN inference | AUC / AUPR | Hurt |
| use too few morphological edges (small K or sigma) (ablation) | KERNEL · hESC GRN inference | AUC / AUPR | Hurt |
| images missing for TFs and genes (reduced morphological completeness) (ablation) | KERNEL · hESC (TFs+500 genes) | AUC / AUPR | Hurt |
| KERNEL with all images missing compared to GENELink (ablation) | KERNEL · hESC (TFs+500 genes) | AUPR | Improved |
| replace static edges (MOTIVE) with dynamic phenotype-based edge augmentation | DTI prediction · CRISPR | AUC: 0.842 -> 0.950 | Improved |
| replace static edges (MOTIVE) with dynamic phenotype-based edge augmentation | DTI prediction · ORF cold-start drug split | AUC: 0.715 -> 0.811 | Improved |
| replace GENELink with KERNEL | GRN inference · hHEP (TFs+1000 genes) | AUPR: 0.312 -> 0.424 | Improved |
| replace GraphSAGE with KERNEL | DTI prediction · CRISPR | Hit@500: 0.611 -> 0.531 | Hurt |
| replace GeSubNet with KERNEL | subtype-specific subnetwork discovery · BRCA | GED: 0.285 -> 0.659 | Improved |
Closest papers by meaning
- GeSubNet: Gene Interaction Inference for Disease Subtype Network Generation (ICLR 2025 · similarity 0.85)
- KINDLE: Knowledge-Guided Distillation for Prior-Free Gene Regulatory Network Inference (NeurIPS 2025 · similarity 0.84)
- OgBench: A Framework for Evaluating Graph Neural Networks on Omics Data (NeurIPS 2026 · similarity 0.84)
- Multi-modal Topology-embedded Graph Learning for Spatially Resolved Genes Prediction from Pathology Images with Prior Gene Similarity Information (CVPR 2025 · similarity 0.84)
- A Cross Modal Knowledge Distillation & Data Augmentation Recipe for Improving Transcriptomics Representations through Morphological Features (ICML 2025 · similarity 0.83)
- Towards the Efficient Inference by Incorporating Automated Computational Phenotypes under Covariate Shift (ICML 2025 · similarity 0.83)
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.
