BACT: nonparametric Bayesian cell typing for single-cell spatial transcriptomics data
2025

BACT: A Bayesian Model for Cell Typing in Spatial Transcriptomics

Sample size: 1207 publication 10 minutes Evidence: high

Author Information

Author(s): Yan Yinqiao, Luo Xiangyu

Primary Institution: Beijing University of Technology

Hypothesis

Can a nonparametric Bayesian model effectively identify cell types in single-cell spatial transcriptomics data without prespecifying the number of cell types?

Conclusion

The BACT model outperforms existing methods in accurately identifying cell types in single-cell spatial transcriptomics data.

Supporting Evidence

  • BACT achieved the highest adjusted Rand index (ARI) of 0.629 in identifying cell types.
  • BACT effectively captured rare cell types and complex spatial distributions.
  • Comparative analysis showed BACT outperformed SpaGCN, STAGATE, BANKSY, and BASS in cell typing accuracy.

Takeaway

BACT is a smart tool that helps scientists figure out what types of cells are in a tissue by looking at their genes and where they are located, without needing to guess how many types there are.

Methodology

BACT uses a nonparametric Bayesian model with a Potts prior to analyze gene expression and spatial data from single-cell transcriptomics.

Potential Biases

Potential biases may arise from the initial assumptions about the number of cell types.

Limitations

The model's execution time is longer compared to other methods due to its iterative sampling process.

Participant Demographics

The study involved various mouse brain tissues and human brain sections.

Statistical Information

P-Value

0.001

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1093/bib/bbae689

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication