Reconstruction Performance Metrics for 3D Spatial Transcriptomics on Three Species

Tianci Song published a 5.5 KB Excel file on figshare in March 2026. The dataset contains performance metrics comparing four reconstruction methods for 3D-projected spatial transcriptomics data. The comparison includes results from human heart, mouse brain, and Drosophila embryo samples.

Use Cases

• Benchmarking reconstruction algorithms based on reported MSE, MAE, and R² metrics
• Comparing the impact of graph Laplacian regularization based on the CTFacTomo variants
• Evaluating method performance across different biological samples based on the human, mouse, and Drosophila data

Strengths

• Provides direct comparison of four specific reconstruction methods (IPF, CTFacTomo, CTFacTomo without graph Laplacian regularization, Tomographer)
• Includes results from three distinct biological samples (human heart, mouse brain, Drosophila embryo)
• Released under a permissive CC-BY-4.0 license
• File size is 5.5 KB, indicating a compact and focused dataset

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download
• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment

Provenance

Source

figshare

Collection Method

Likely contains computational results from algorithm benchmarking.

Time Range

Publication date is 2026-03-24.

Freshness

Last updated 2026-03 24 20:09:35

Geography

Data is not inherently geographic; it relates to biological samples from human, mouse, and Drosophila.