DestVI

DestVI models continuous cell-state variation to deconvolve spatial transcriptomics data and estimate cell-type-specific gene expression at sub-spot resolution.

Key Features:

• Multi-Resolution Analysis: Enables examination of gene expression at multiple spatial resolutions to resolve fine-grained cell-state variation.
• Integration with scRNA-seq: Integrates spatial transcriptomics data with single-cell RNA sequencing (scRNA-seq) to inform cell-type and state deconvolution.
• Probabilistic Modeling: Employs probabilistic modeling and variational inference to model continuous variation within cell types.
• Higher-Resolution Estimation: Demonstrates via simulations the ability to provide higher-resolution cell-type-specific gene expression estimates compared to existing methods.
• Automated Analysis Pipeline: Provides an automated pipeline for analyzing single tissue slices and comparing tissues across conditions.

Scientific Applications:

• Immune Crosstalk Analysis: Characterized spatial organization and interactions of immune cells in lymph nodes responding to infection.
• Tumor Microenvironment Exploration: Mapped cellular organization and identified condition-specific gene expression changes in a mouse tumor model across tissue regions.

Methodology:

Integrates spatial transcriptomics with single-cell RNA sequencing, models continuous variation within cell types using probabilistic approaches fitted by variational inference, and employs simulations for performance evaluation.