Central-to-Local Adaptive Generative Diffusion Framework for Improving Gene Expression Prediction in Data-Limited Spatial Transcriptomics

arXiv cs.LG / 3/31/2026

💬 OpinionIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces C2L-ST, a central-to-local adaptive generative diffusion framework designed to improve gene expression prediction from spatial transcriptomics under severe data scarcity.
It pretrains a global central diffusion model on large histopathology datasets to learn transferable morphological priors, then adapts institution-specific local models using lightweight gene-conditioned modulation from a small set of paired image–gene spots.
The method generates synthetic histology patches that are both visually/structurally realistic and molecularly consistent, demonstrating strong overlap in embeddings with real data across multiple organs.
When synthetic image–gene pairs are used to augment downstream training, gene expression prediction accuracy and spatial coherence improve to near-real-data performance while using only a fraction of sampled spots.
Overall, C2L-ST is presented as a scalable, data-efficient approach for molecular-level data augmentation that better integrates histology and transcriptomics in spatial biology.

Abstract

Spatial Transcriptomics (ST) provides spatially resolved gene expression profiles within intact tissue architecture, enabling molecular analysis in histological context. However, the high cost, limited throughput, and restricted data sharing of ST experiments result in severe data scarcity, constraining the development of robust computational models. To address this limitation, we present a Central-to-Local adaptive generative diffusion framework for ST (C2L-ST) that integrates large-scale morphological priors with limited molecular guidance. A global central model is first pretrained on extensive histopathology datasets to learn transferable morphological representations, and institution-specific local models are then adapted through lightweight gene-conditioned modulation using a small number of paired image-gene spots. This strategy enables the synthesis of realistic and molecularly consistent histology patches under data-limited conditions. The generated images exhibit high visual and structural fidelity, reproduce cellular composition, and show strong embedding overlap with real data across multiple organs, reflecting both realism and diversity. When incorporated into downstream training, synthetic image-gene pairs improve gene expression prediction accuracy and spatial coherence, achieving performance comparable to real data while requiring only a fraction of sampled spots. C2L-ST provides a scalable and data-efficient framework for molecular-level data augmentation, offering a domain-adaptive and generalizable approach for integrating histology and transcriptomics in spatial biology and related fields.