From an H&E slide to gene expression, in minutes.
Precision oncology at your fingertips.
Upload a whole slide image. The published Path2Omics model infers 5,000–8,000 genes per cancer type, runs pathway enrichment (ssGSEA), and generates a plain-English AI summary of results.
For research use only · not for clinical decision-making
Sample Output
Inferred| Gene | Expression | Z-score |
|---|---|---|
| TP53 | 12.4 | +2.31 |
| EGFR | 8.7 | +1.45 |
| MYC | 6.2 | -0.82 |
| BRCA1 | 4.9 | -1.67 |
| PTEN | 3.1 | -2.14 |
Example gene expression inference from H&E WSI
How the pipeline works
A four-stage workflow from raw histology to interpretable biology.
Upload & Annotate
Upload H&E whole slide images (single or batch), annotate cancer type, and optionally add clinical metadata.
AI Inference
Deep learning model analyzes morphology patterns to infer per-gene expression levels without sequencing.
ssGSEA & Networks
Automated pathway enrichment analysis and key gene network interactions to uncover biological insights.
AI Summary
Claude generates a plain-English narrative summarizing the most significant findings from your analysis.
What you provide
Simple inputs, powerful insights. Start with just your slide images.
H&E Whole Slide Images
RequiredUpload full-resolution WSI files. Supports single samples or batch uploads for multiple slides.
Cancer type annotation
RequiredSpecify the cancer type for each submitted sample to enable accurate model inference.
Clinical metadata
OptionalOptionally provide phenotypic data like survival times or drug response for association analysis.
TCGA cohort selection
OptionalOptionally select from curated TCGA cohorts for comparative analysis or exploration.
Comprehensive analysis outputs
Everything you need to interpret your H&E slides at the molecular level.
Gene expression profiles
Sortable, filterable table of inferred per-gene expression with statistical significance metrics.
Pathway enrichment
Single-sample GSEA with cancer hallmarks and curated pathways. Highlights active and suppressed programs.
Narrative summary
AI-generated paragraph summarizing the most salient findings, powered by Claude.
TCGA cohorts
Optional access to curated TCGA sample cohorts for comparative analysis and exploration.
Phenotype associations
Correlate gene expression with clinical metadata like survival outcomes and drug response.
Privacy controls
Opt-in data retention consent. Your data is handled securely with usage metrics tracked transparently.
Research use only
HistoOmics-AI is intended for academic and non-profit research purposes only. Results must not be used for clinical decision-making or patient treatment. All uploaded images must be de-identified by the user before submission. By using this tool, you agree to these terms and acknowledge that outputs are for research exploration only.
Curious what your slide reveals?
Try the public demo with sample data, or create a free account to upload your own H&E whole slide images.