sc3D: A Comprehensive Tool for 3D Spatial Transcriptomic Analysis.

Area of Science:

• Genomics
• Bioinformatics
• Developmental Biology

Background:

• Serial spatial omics technologies provide 2D gene expression data but lack 3D continuity.
• Reconstructing 3D volumes is crucial for understanding molecular organization in complex biological structures like organs and embryos.

Purpose of the Study:

• To introduce sc3D, an open-source Python framework for reconstructing and analyzing volumetric spatial omics data.
• To enable the visualization and analysis of gene expression patterns in three dimensions from serial 2D tissue sections.

Main Methods:

• sc3D registers consecutive spatial transcriptomic sections and interpolates bead coordinates in 3D.
• The framework integrates slice alignment, 3D reconstruction, optional downsampling, and visualization using a napari plugin.
• Data is stored in an AnnData object compatible with Scanpy for downstream analysis.

Main Results:

• sc3D successfully reconstructed continuous tissue morphologies from Slide-seq and Stereo-seq datasets of mouse embryos (E8.5 and E16.5).
• The framework recovered cardiac anatomical markers and validated anterior-posterior gradients of Hox gene expression.
• Interactive 3D visualization and spatial differential gene expression analysis were demonstrated.

Conclusions:

• sc3D facilitates reproducible 3D reconstruction and analysis of spatial omics data across different samples and platforms.
• The tool enhances the study of tissue architecture and molecular gradients in developmental contexts.
• sc3D supports virtual in situ hybridization and spatial differential expression analysis for comprehensive biological insights.