Gene-DML: Dual-Pathway Multi-Level Discrimination for Gene Expression Prediction from Histopathology Images

TL;DR

Gene-DML introduces a dual-pathway multi-level discrimination framework that significantly improves gene expression prediction from histopathology images by better aligning morphological and transcriptional data across multiple scales.

Contribution

It presents a novel unified approach that structures latent space with dual-pathway discrimination, capturing multi-scale and cross-level relationships for enhanced cross-modal alignment.

Findings

Achieves state-of-the-art gene expression prediction accuracy.

Effectively captures multi-scale morphological-transcriptional relationships.

Demonstrates robust generalization across diverse datasets.

Abstract

Accurately predicting gene expression from histopathology images offers a scalable and non-invasive approach to molecular profiling, with significant implications for precision medicine and computational pathology. However, existing methods often underutilize the cross-modal representation alignment between histopathology images and gene expression profiles across multiple representational levels, thereby limiting their prediction performance. To address this, we propose Gene-DML, a unified framework that structures latent space through Dual-pathway Multi-Level discrimination to enhance correspondence between morphological and transcriptional modalities. The multi-scale instance-level discrimination pathway aligns hierarchical histopathology representations extracted at local, neighbor, and global levels with gene expression profiles, capturing scale-aware morphological-transcriptional relationships. In parallel, the cross-level instance-group discrimination pathway enforces structural consistency between individual (image/gene) instances and modality-crossed (gene/image, respectively) groups, strengthening the alignment across modalities. By jointly modeling fine-grained and structural-level discrimination, Gene-DML is able to learn robust cross-modal representations, enhancing both predictive accuracy and generalization across diverse biological contexts. Extensive experiments on public spatial transcriptomics datasets demonstrate that Gene-DML achieves state-of-the-art performance in gene expression prediction. The code and processed datasets are available at https://github.com/YXSong000/Gene-DML.