Related Experiment Video
Updated: Jun 5, 2026
10:22
Comprehensive Spatial Profiling of Species-agnostic Transcriptomes via Stereo-seq
Published on: October 31, 2025
Atlas-Level Single-Cell and Spatial Transcriptomics Data Integration via PRIME
Xinchao Wu1, Xuefei Wang2, Jieqiong Wang3
1Department of Genetics, Cell Biology, and Anatomy, University of Nebraska Medical Center, Omaha, NE.
Biorxiv : the Preprint Server for Biology
|June 4, 2026
Summary
PRIME is a new computational framework for integrating large-scale single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics (ST) data. It effectively reconciles batch effects while preserving biological insights for cell atlases.
Area of Science:
- Computational Biology
- Genomics
- Bioinformatics
Background:
- Atlas-scale cellular cartography using scRNA-seq and ST is crucial for understanding cell identity and disease mechanisms.
- Integrating heterogeneous data from multiple sources, donors, and technologies presents significant computational challenges, especially with imbalanced cell-type compositions.
- Existing integration methods often struggle to balance batch effect correction with the preservation of subtle biological variations and spatial information.
Purpose of the Study:
- To introduce PRIME (Projection-based Robust Integration via Manifold Embedding), an ensemble framework for robust atlas-level integration of scRNA-seq and ST data.
- To develop a method that can jointly reconcile complex batch effects across numerous datasets while preserving biological structure and accommodating spatial coordinates.
- To provide a computationally tractable solution for integrating millions of cells.
Main Methods:
- PRIME employs an ensemble approach combining random-projection-based consensus anchoring and graph-Laplacian correction.
- Consensus voting across random projections identifies reliable cell-pair anchors, reducing projection-specific distortions.
- For ST data, PRIME integrates expression-based anchors with spatial coordinates using a graph-Laplacian objective for simultaneous alignment and spatial coherence.
Main Results:
- PRIME consistently outperforms state-of-the-art methods in batch correction and biological conservation for both scRNA-seq and ST integration.
- The framework successfully preserves developmental trajectories in human hematopoiesis and cortical laminar architecture in spatial transcriptomics data.
- PRIME accurately recovers drug-target relationships in a large perturbation atlas (>1 million cells) while mitigating batch effects.
Conclusions:
- PRIME offers a versatile and scalable solution for integrating large-scale scRNA-seq and ST datasets.
- The framework demonstrates superior performance in preserving biological accuracy and spatial information across diverse applications.
- PRIME advances the field of single-cell and spatial data integration, enabling more comprehensive cellular atlases and deeper biological insights.
