FLAG redefines spatial gene expression prediction as structured distribution modeling via latent diffusion, integrating a spatial graph encoder and GFM alignment to capture gene-gene and spatial relationships, with new GSC and SSC metrics showing improved structural fidelity over pointwise baselines
Note that since CellPLM produces cell-level embeddings, we apply average pooling along the gene dimensionto align its output with our target gene embedding space
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FLAG: Foundation model representation with Latent diffusion Alignment via Graph for spatial gene expression prediction
FLAG redefines spatial gene expression prediction as structured distribution modeling via latent diffusion, integrating a spatial graph encoder and GFM alignment to capture gene-gene and spatial relationships, with new GSC and SSC metrics showing improved structural fidelity over pointwise baselines