Generative 3D Spatial Reconstruction of Cellular Organization from Single Cell Expression
- 1UNIVERSITY OF CENTRAL FLORIDA
Abstract
Single-cell RNA sequencing profiles cellular states at scale but removes their three-dimensional (3D) spatial organization. Reconstructing this organization requires recovering not only individual cell positions but also the physical size, geometry, and neighborhoods of an entire tissue. Existing methods either map cells onto measured spatial references or reconstruct two-dimensional sections independently, leaving the joint reconstruction of a new 3D specimen unresolved. Moreover, predicting the average plausible position of each cell can reduce per-cell error while compressing tissue geometry. We introduce ORBIS, a generative framework that jointly reconstructs whole 3D specimens from single-cell gene expression without requiring a spatial reference at inference. ORBIS learns spatial organization through flow matching in a shared physical coordinate frame, using training views sampled from sparse serial sections, tissue blocks, and whole specimens. At inference, a shared spatial state is iteratively updated through dynamically regrouped windows, enabling the generation of millions of cell positions without independently reconstructing and stitching tissue regions. We evaluate physical size, cell-population shape, per-cell placement, and neighborhood organization separately to distinguish accurate tissue geometry from low per-cell error. Across eight held-out specimens spanning whole mouse brains, thick brain blocks, and Drosophila embryos, ORBIS achieves lower cell-population shape error than all five evaluated baselines on every test and reconstructs specimens containing up to 3.7 million cells. On a single GPU, ORBIS samples million-cell brains in 0.9-2.8 minutes, approximately 38 times faster than the sampler of LUNA-3D, our 3D extension of the state-of-the-art generative spatial reconstruction method LUNA. On two held-out brain tests, sampling reduces size error by 3.3-fold and shape error by 1.8-fold compared with a single-call variant. On an independent dissociated single-cell brain atlas, ORBIS returns on average 37.9% of each region's cells to their dissection region across 23 regions, compared with 28.5% for LUNA-3D.
Related
- Quantitative Head-to-Tail Analysis Reveals Heterogeneous Cellular Organization and Connexin Expression in the Porcine Sinoatrial Node
- Spatial organization of voltage-gated ion channel expression across molecularly defined neuronal populations in the mouse mammillary bodies
- Deciphering Spatial Cellular Organization through RNA-Anchored Multi-Omics Deconvolution
- SRF3 couples cellular iron homeostasis and defense responses through ubiquitin-dependent nano-organization
- Speciesformer learns conserved cellular states for cross-species generative virtual cell modeling
BibTeX
@misc{azim2026generative,
title = {{Generative 3D Spatial Reconstruction of Cellular Organization from Single Cell Expression}},
author = {E. Azim and M. N. Alif and T. H. Hwang and H. Zheng and W. i. Zhang},
howpublished = {bioRxiv},
year = {2026},
month = nov,
doi = {10.64898/2026.11.05.764656},
url = {https://www.biorxiv.org/content/10.64898/2026.11.05.764656v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Generative 3D Spatial Reconstruction of Cellular Organization from Single Cell Expression
AU - E. Azim
AU - M. N. Alif
AU - T. H. Hwang
AU - H. Zheng
AU - W. i. Zhang
PY - 2026
DA - 2026/11/05
PB - bioRxiv
DO - 10.64898/2026.11.05.764656
UR - https://www.biorxiv.org/content/10.64898/2026.11.05.764656v1
AB - Single-cell RNA sequencing profiles cellular states at scale but removes their three-dimensional (3D) spatial organization. Reconstructing this organization requires recovering not only individual cell positions but also the physical size, geometry, and neighborhoods of an entire tissue. Existing methods either map cells onto measured spatial references or reconstruct two-dimensional sections independently, leaving the joint reconstruction of a new 3D specimen unresolved. Moreover, predicting the average plausible position of each cell can reduce per-cell error while compressing tissue geometry. We introduce ORBIS, a generative framework that jointly reconstructs whole 3D specimens from single-cell gene expression without requiring a spatial reference at inference. ORBIS learns spatial organization through flow matching in a shared physical coordinate frame, using training views sampled from sparse serial sections, tissue blocks, and whole specimens. At inference, a shared spatial state is iteratively updated through dynamically regrouped windows, enabling the generation of millions of cell positions without independently reconstructing and stitching tissue regions. We evaluate physical size, cell-population shape, per-cell placement, and neighborhood organization separately to distinguish accurate tissue geometry from low per-cell error. Across eight held-out specimens spanning whole mouse brains, thick brain blocks, and Drosophila embryos, ORBIS achieves lower cell-population shape error than all five evaluated baselines on every test and reconstructs specimens containing up to 3.7 million cells. On a single GPU, ORBIS samples million-cell brains in 0.9-2.8 minutes, approximately 38 times faster than the sampler of LUNA-3D, our 3D extension of the state-of-the-art generative spatial reconstruction method LUNA. On two held-out brain tests, sampling reduces size error by 3.3-fold and shape error by 1.8-fold compared with a single-call variant. On an independent dissociated single-cell brain atlas, ORBIS returns on average 37.9% of each region's cells to their dissection region across 23 regions, compared with 28.5% for LUNA-3D.
ER -