arXiv:2603.19766v1 Announce Type: new 
Abstract: Spatial transcriptomics (ST) enables spot-level in situ expression profiling, but its high cost and limited throughput motivate predicting expression directly from HE-stained histology. Recent advances explore using score- or flow-based generative models to estimate the conditional distribution of gene expression from histology, offering a flexible alternative to deterministic regression approaches. However, most existing generative approaches omit explicit modeling of gene-gene dependencies, undermining biological coherence. Single-cell foundation models (sc-FMs), pre-trained across diverse cell populations, capture these critical gene relationships that histology alone cannot reveal. Yet, applying expression-only sc-FMs to histology-conditioned expression modeling is nontrivial due to the absence of a visual pathway, a mismatch between their pre-training and conditional ST objectives, and the scarcity of mixed-cell ST supervision. To address these challenges, we propose HINGE (HIstology-coNditioned GEneration), which retrofits a pre-trained sc-FM into a conditional expression generator while mostly preserving its learned gene relationships. We achieve this by introducing SoftAdaLN, a lightweight, identity-initialized modulation that injects layer-wise visual context into the backbone, coupled with an expression-space masked diffusion objective and a warm-start curriculum to ensure objective alignment and training stability. Evaluated on three ST datasets, ours outperforms state-of-the-art baselines on mean Pearson correlation and yields more accurate spatial marker expression patterns and higher pairwise co-expression consistency, establishing a practical route to adapt pre-trained sc-FMs for histology-conditioned spatial expression generation.

قدمت دراسة جديدة طريقة لتوليد ملفات تعريف التعبير الجيني المكاني مباشرة من صور الأنسجة باستخدام نموذج أساسي للخلايا الفردية (sc-FM) تم تدريبه مسبقًا. تتناول هذه الطريقة قيود النسخ الجيني المكاني التقليدي، الذي غالبًا ما يكون مكلفًا ويعاني من انخفاض الإنتاجية، من خلال الاستفادة من النماذج التوليدية التي يمكن أن تقدر توزيعات التعبير الجيني دون الاعتماد فقط على طرق الانحدار الحتمي.

Un nuevo estudio ha introducido un método para generar perfiles de expresión génica espacial directamente a partir de imágenes de histología utilizando un modelo de fundación de célula única (sc-FM) preentrenado. Este enfoque aborda las limitaciones de la transcriptómica espacial tradicional, que a menudo es costosa y tiene un bajo rendimiento, al aprovechar modelos generativos que pueden estimar distribuciones de expresión génica sin depender únicamente de métodos de regresión deterministas.

Une nouvelle étude a présenté une méthode pour générer des profils d'expression génique spatiale directement à partir d'images d'histologie en utilisant un modèle de fondation à cellule unique pré-entraîné (sc-FM). Cette approche répond aux limitations de la transcriptomique spatiale traditionnelle, souvent coûteuse et à faible débit, en s'appuyant sur des modèles génératifs capables d'estimer les distributions d'expression génique sans se fier uniquement à des méthodes de régression déterministes.

A new study has introduced a method for generating spatial gene expression profiles directly from histology images using a pre-trained single-cell foundation model (sc-FM). This approach addresses the limitations of traditional spatial transcriptomics, which is often costly and has low throughput, by leveraging generative models that can estimate gene expression distributions without relying solely on deterministic regression methods.

Adapting a Pre-trained Single-Cell Foundation Model to Spatial Gene Expression Generation from Histology Images

Useful picks

LucidQuery AI

FETCH HIVE

Ask On Data

X Headshot

Dyad

Dynamiq