AI to map tissue structure and gene activity in 3D
Novel geometric deep learning models for tissue structure-aware spatial expression representations from spatially resolved single-cell transcriptomics data
This project builds AI models that read 2D and 3D maps of gene activity in tissues to help spot disease patterns, including in lungs affected by COVID-19.
Quick facts
| Grant type | R21 grant |
|---|---|
| Study type | NIH-funded research |
| Funding institution | University of Pittsburgh at Pittsburgh NIH-funded |
| Lab location | 1 site (Pittsburgh, United States) |
| Project ID | NIH-11174561 on NIH RePORTER |
What this research studies
From a patient's perspective, researchers will create new geometric deep learning tools that learn from maps showing where thousands of genes sit inside intact tissues. They will work with spatially resolved single-cell technologies such as in‑situ sequencing and barcode-based hybridization that preserve the exact 2D and 3D locations of transcripts. The team will train models to recognize tissue structures and molecular signatures of pathology using datasets that include normal and COVID‑19 patient lung samples. If successful, these tools could reveal disease patterns and molecular changes that current methods miss.
Who could benefit from this research
Good fit: Ideal candidates are people who can provide tissue samples (for example lung tissue from biopsy or surgery, including COVID‑19 patient samples) or who can share their tissue data for research.
Not a fit: People seeking immediate treatment or those who cannot or will not provide tissue samples or share data are unlikely to receive direct benefit from this computational methods project.
Why it matters
Potential benefit: If successful, this work could make it easier to detect and understand tissue-level disease changes and guide development of better diagnostics and treatments.
How similar studies have performed: Related spatial transcriptomics and AI approaches exist, but current graph-based methods have struggled to capture tissue histology, so this geometric deep learning approach is relatively novel and not yet proven clinically.
Where this research is happening
Pittsburgh, United States
- University of Pittsburgh at Pittsburgh — Pittsburgh, United States (Active)
Researchers
- Principal investigator: Huang, Yufei — University of Pittsburgh at Pittsburgh
- Study coordinator: Huang, Yufei
About this research
- This is an active NIH-funded research project — typically early-stage science, not a clinical trial accepting patient enrollment.
- Some NIH-funded labs run parallel clinical studies or seek volunteers for related work. To check, contact the principal investigator or institution listed above.
- For full project details, budget, and progress reports, visit the official NIH RePORTER page below.