GenVinci reconstructs 3D molecular tissue scenes from histology to map senescence across aging and disease

Spatial omics has generated measurements at unprecedented scale, yet computational approaches remain largely confined to individual modalities and 2D representations. Here we assemble ImageOmicsNet-300M, a corpus of 300 million single-cell and spatial transcriptomic profiles, including 24.9 million H&E-predicted virtual spatial transcriptomics profiles, and 4,623 H&E sections. Of these H&E sections, we generated 1,782 from aging mouse organs and serially sectioned human skin spanning five decades, with matched spatial transcriptomics and snRNA-seq. Building on this resource, we develop GenVinci, a multimodal tissue foundation model that reconstructs 3D tissue scenes from routine 2D histology. GenVinci predicts spatial gene expression from H&E with zero-shot generalization and maps senescence-associated states across 85.8 million cells, revealing cell-type-specific patterns and disease-associated remodeling. Integration with human genetics implicates suprabasal keratinocytes in atopic dermatitis and keratinocyte carcinomas, linked to higher and lower p16 and p21 expression, respectively. These findings establish GenVinci as a framework mapping senescence in 3D across scales.

Authors

Institutions

Publication Details

Journal
bioRxiv (Cold Spring Harbor Laboratory)
Published
2026-11-18
DOI
https://doi.org/10.64898/2026.11.17.767412
Primary Topic
Single-cell and spatial transcriptomics
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

GenVinci reconstructs 3D molecular tissue scenes from histology to map senescence across aging and disease

Aviv Regev, SU Ying-li, Collin M. Stultz, Shadmehr Demehri et al.
3 links
bioRxiv (Cold Spring Harbor Laboratory)
Single-cell and spatial transcriptomics
preprint

GenVinci reconstructs 3D molecular tissue scenes from histology to map senescence across aging and disease

Aviv Regev, SU Ying-li, Collin M. Stultz, Shadmehr Demehri, Peter T. C. So, Zhixun Dou, Xuemeng Li, Jonathan C. Silverstein, Stefano Sol, Wei Min, Francesco Monticolo, Denny Sakkas, Christine Guo Lian, Anna Mandinova, Ke Zhang, Jian Shu, Judy Zhou, Kristina A. Todorova, Xingjian Chen, Ryota Tanaka, Hemali P. Phatnani, Yanwan Cao, Yinghan Wu, Haochun Huang, Styliani Sapantzi, Jianhuan Qi, Ziegler, Justin, author, Charles Comiter, Jeon Woong Kang, Max Condouret, Heng Yang, Yanran Zhou
preprint en

Abstract

Spatial omics has generated measurements at unprecedented scale, yet computational approaches remain largely confined to individual modalities and 2D representations. Here we assemble ImageOmicsNet-300M, a corpus of 300 million single-cell and spatial transcriptomic profiles, including 24.9 million H&E-predicted virtual spatial transcriptomics profiles, and 4,623 H&E sections. Of these H&E sections, we generated 1,782 from aging mouse organs and serially sectioned human skin spanning five decades, with matched spatial transcriptomics and snRNA-seq. Building on this resource, we develop GenVinci, a multimodal tissue foundation model that reconstructs 3D tissue scenes from routine 2D histology. GenVinci predicts spatial gene expression from H&E with zero-shot generalization and maps senescence-associated states across 85.8 million cells, revealing cell-type-specific patterns and disease-associated remodeling. Integration with human genetics implicates suprabasal keratinocytes in atopic dermatitis and keratinocyte carcinomas, linked to higher and lower p16 and p21 expression, respectively. These findings establish GenVinci as a framework mapping senescence in 3D across scales.

bioRxiv (Cold Spring Harbor Laboratory)
Broad Institute (US), Brigham and Women's Hospital (US), Harvard University (US), University of Pittsburgh (US), Genentech (US), Massachusetts General Hospital (US), Harvard–MIT Division of Health Sciences and Technology (US), Center for Cancer Research (US), New York Genome Center (US), Harvard Stem Cell Institute (US), Mass General Brigham (US), Massachusetts Institute of Technology (US), Columbia University (US)
Massachusetts General Hospital, Harvard University, Carnegie Mellon University, Massachusetts Life Sciences Center, National Institutes of Health, FAS Division of Science, Harvard University, Common Fund
Good health and well-being
Single-cell and spatial transcriptomics
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon