ST2HE: Enhancing spatial transcriptomics interpretability via virtual staining for histological annotation

This repository contains the code for ST2HE, a cross-platform generative framework that synthesizes virtual hematoxylin and eosin (H&E) images directly from high-resolution spatial transcriptomics (HR-ST) data. ST2HE integrates nuclei morphology and spatial transcript coordinates using a one-step diffusion model, enabling histologically informative image generation across diverse tissue types and HR-ST platforms.

The repository includes:

• ST2HE model weights for ST2HE-CondGen and ST2HE-UnCondGen variants
• Scripts for input data preparation including DAPI/H&E image registration and transcript overlay generation
• Training scripts for model fine-tuning on new tissue types
• Inference scripts for virtual H&E generation

This code accompanies the manuscript: "ST2HE: Enhancing spatial transcriptomics interpretability via virtual staining for histological annotation", published in Briefings in Bioinformatics (2026).