DANST enables cell-type deconvolution in spatial transcriptomics using deep domain adversarial neural networks.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Spatial transcriptomics enables gene expression analysis within tissue context.
• Accurate cell type deconvolution from spatial data is crucial for biological insights.
• Existing methods face challenges in precisely identifying cell proportions in complex tissues.

Purpose of the Study:

• To introduce DANST, a deep domain adversarial neural network framework for accurate cell type deconvolution in spatial transcriptomics.
• To leverage single-cell RNA sequencing (scRNA-seq) and inferred spatial coordinates for improved deconvolution.
• To enhance the analysis of the tumor microenvironment and explore clinical utility.

Main Methods:

• Integration of scRNA-seq with inferred spatial coordinates to create pseudo-spatial data.
• Utilization of a variational autoencoder for refined feature representation learning.
• Implementation of a domain adversarial architecture to align pseudo and real spatial data distributions for accurate label transfer.

Main Results:

• DANST demonstrates superior deconvolution accuracy compared to existing methods on human and mouse datasets.
• The framework effectively learns feature representations and aligns data distributions.
• Successful application in analyzing tumor microenvironment composition.

Conclusions:

• DANST provides a robust and accurate solution for cell type deconvolution in spatial transcriptomics.
• The method shows significant potential for advancing tumor microenvironment research.
• DANST's effectiveness suggests broad applicability in clinical settings for spatial biology analysis.