Research map / Probabilistic and scientific modeling / Molecules and proteins
TRIDENT: A Trimodal Cascade Generative Framework for Drug and RNA-Conditioned Cellular Morphology Synthesis
Rui Peng, Ziru Liu, Lingyuan Ye, Yuxing Lu, Boxin Shi, Jinzhuo Wang
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 RNA (pre-perturbation gene expression) conditioning alongside drug embedding (ablation) | TRIDENT · MorphoGene in-distribution test | FID: 115.770 -> 49.770 | Improved |
| add RNA conditioning alongside drug embedding (ablation) | TRIDENT · MorphoGene in-distribution test | KID: 0.132 -> 0.013 | Improved |
| add RNA conditioning alongside drug embedding (ablation) | TRIDENT · MorphoGene out-of-distribution test | FID: 194.239 -> 126.150 | Improved |
| add RNA conditioning alongside drug embedding (ablation) | TRIDENT · MorphoGene out-of-distribution test | KID: 0.293 -> 0.222 | Improved |
| add RNA conditioning alongside drug embedding (ablation) | TRIDENT · MorphoGene in-distribution test | IS: 2.381 -> 2.240 | Improved |
| use TRIDENT (drug + RNA conditioned latent diffusion) instead of MorphoDiff | conditional morphology synthesis · MorphoGene in-distribution test | FID: 250.290 -> 49.770 | Improved |
| use TRIDENT instead of MorphoDiff on unseen compounds | conditional morphology synthesis · MorphoGene out-of-distribution test | FID: 387.135 -> 126.150 | Improved |
| use TRIDENT instead of unconditional Stable Diffusion | conditional morphology synthesis · MorphoGene in-distribution test | KID: 0.378 -> 0.013 | Improved |
| train jointly on VAE loss predicting post-perturbation RNA together with latent diffusion loss | TRIDENT conditioning module · MorphoGene (44 ID compounds) | Pearson correlation of predicted LFC: 0.957 | Improved |
Closest papers by meaning
- Cradle-VAE: Enhancing Single-Cell Gene Perturbation Modeling with Counterfactual Reasoning-based Artifact Disentanglement (AAAI 2025 · similarity 0.85)
- STRAND: Sequence-Conditioned Transport for Single-Cell Perturbations (NeurIPS 2026 · similarity 0.85)
- A Cross Modal Knowledge Distillation & Data Augmentation Recipe for Improving Transcriptomics Representations through Morphological Features (ICML 2025 · similarity 0.84)
- Learning Cross-Domain Representations for Transferable Drug Perturbations on Single-Cell Transcriptional Responses (AAAI 2025 · similarity 0.84)
- CellFlux: Simulating Cellular Morphology Changes via Flow Matching (ICML 2025 · similarity 0.83)
- MorphoHELM: A Comprehensive Benchmark for Evaluating Representations for Microscopy-Based Morphology Assays (NeurIPS 2026 · 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.
