Research map / Probabilistic and scientific modeling / Molecules and proteins
MoLF: Mixture-of-Latent-Flow for Pan-Cancer Spatial Gene Expression Prediction from Histology
Susu Hu, Stefanie Speidel
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 |
|---|---|---|---|
| replace dense attention velocity network with Mixture-of-Experts (ablation) | MoLF · HEST-1k Top50 HVG average | PCC: 0.209 -> 0.383 | Improved |
| replace dense attention velocity network with Mixture-of-Experts (ablation) | MoLF · HEST-1k Hallmark genes overall | PCC: 0.127 -> 0.185 | Improved |
| use MoE instead of dense transformer in synthetic task (ablation) | MoLF flow · Conditional eight Gaussians | 2-Wasserstein distance: 0.315 -> 0.292 | Improved |
| add sinusoidal positional encoding on image features (ablation) | MoLF · HEST-1k Top50 HVG average | PCC: 0.386 -> 0.383 | No clear effect |
| add sinusoidal positional encoding on image features (ablation) | MoLF · HEST-1k Hallmark genes overall | PCC: 0.173 -> 0.185 | Improved |
| train on pan-cancer data instead of skin cancer only (ablation) | MoLF · Zero-shot mouse melanoma | PCC: 0.127 -> 0.202 | Improved |
| MoLF flow matching with MoE versus foundation model STPath | MoLF · HEST-1k Top50 HVG average | PCC: 0.361 -> 0.382 | Improved |
| MoLF versus STFlow single-flow baseline | flow matching for ST · HEST-1k Hallmark genes overall | PCC: 0.122 -> 0.185 | Improved |
| use the large foundation model STPath instead of an MLP | Hallmark gene prediction · HEST-1k Hallmark genes overall | PCC: 0.171 -> 0.143 | Hurt |
| use diffusion (STEM) instead of an MLP | HVG prediction · HEST-1k Top50 HVG average | PCC: 0.341 -> 0.196 | Hurt |
| MoLF versus STPath on zero-shot cross-species data | MoLF · Zero-shot mouse melanoma | PCC: 0.090 -> 0.202 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- STFlow — STFlow: Data-Coupled Flow Matching for Geometric Trajectory Simulation (ICML 2026) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Scalable Generation of Spatial Transcriptomics from Histology Images via Whole-Slide Flow Matching (ICML 2025 · similarity 0.89)
- Diffusion Generative Modeling for Spatially Resolved Gene Expression Inference from Histology Images (ICLR 2025 · similarity 0.89)
- HEST-1k: A Dataset For Spatial Transcriptomics and Histology Image Analysis (NeurIPS 2024 · similarity 0.88)
- MERGE: Multi-faceted Hierarchical Graph-based GNN for Gene Expression Prediction from Whole Slide Histopathology Images (CVPR 2025 · similarity 0.88)
- STimage-1K4M: A histopathology image-gene expression dataset for spatial transcriptomics (NeurIPS 2024 · similarity 0.88)
- GeneFlow: Translation of Single-cell Gene Expression to Histopathological Images via Rectified Flow (NeurIPS 2025 · similarity 0.88)
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.
