Turning Routine Slides Into Molecular Maps
Could a standard H&E slide contain enough information to predict gene expression, methylation patterns, and even aspects of spatial biology?
-
By
-
Helen Bristow
-
July 13, 2026
-
-
1
Researchers developed two AI platforms, Path2Omics and Path2Space, to enhance molecular insights from pathology slides.
-
2
Path2Omics predicts bulk molecular information from pathology images, focusing on transcriptomics and methylation across 30 tumor types.
-
3
Path2Space aims to infer spatial transcriptomics from digital pathology images, providing detailed insights into tumor microenvironments.
-
4
AI models can generate predictions from standard H&E slides, outperforming traditional multi-omics approaches in predicting treatment response.
-
5
The AI platforms aim to make molecular insights more accessible, reducing costs and turnaround times compared to conventional sequencing.
