RNA-FM: Flow-Matching Generative Model for Genome-wide RNA-Seq Prediction

TL;DR

RNA-FM introduces a flow-matching generative model that predicts genome-wide RNA expression from histopathology images, capturing biological heterogeneity and providing interpretable results.

Contribution

It presents a novel flow-matching framework that formulates transcriptomic prediction as a continuous transport problem, improving accuracy and interpretability over existing methods.

Findings

RNA-FM outperforms state-of-the-art models in RNA prediction accuracy.

The model captures biological heterogeneity and uncertainty.

It enables scalable, interpretable gene expression imputation.

Abstract

Histopathology whole-slide images (WSIs) are routinely acquired in clinical practice and contain rich tissue morphology but lack direct molecular architecture and functional programs defining pathological states, whereas RNA sequencing (RNA-seq) provides genome-wide transcriptional profiles at substantial cost, thereby motivating WSI-based genome-wide transcriptomic prediction. Existing approaches for predicting gene expression from WSIs predominantly rely on deterministic regression with one-to-one mapping, limiting their ability to capture biological heterogeneity and predictive uncertainty. We propose RNA-FM, a flow-matching generative framework for genome-wide bulk RNA-seq prediction from WSIs. RNA-FM formulates transcriptomic prediction as a continuous-time conditional transport problem, learning a velocity field that maps a simple prior to the target gene expression distribution conditioned on morphologies. By integrating pathway-level structure, RNA-FM enables scalable and biologically interpretable genome-wide gene expression imputation. Extensive experiments demonstrate that RNA-FM consistently outperforms state-of-the-art approaches while maintaining biological meaningfulness. Code is available at https://github.com/YXSong000/RNA-FM.