RCTD

RCTD deconvolves spatial transcriptomics measurements to estimate cell-type composition using single-cell RNA sequencing reference profiles.

Key Features:

• Decomposition of Cell Type Mixtures: Uses profiles derived from single-cell RNA sequencing to deconvolute spatial transcriptomic data and estimate individual cell-type contributions within mixed signals.
• Accounting for Technical Variability: Mitigates platform effects and systematic technical variability across sequencing modalities to improve robustness across datasets.
• Improved Cell Type Assignment: Enhances cell-type assignment accuracy and reproduces known localization patterns on Slide-seq data, including cell types and subtypes in the cerebellum and hippocampus.
• Detection of Mixtures and Identification of Cell Types: Detects mixed signals and identifies distinct cell types within assessment datasets.
• Discovery of Environment-Dependent Gene Expression: Enables discovery of genes whose expression is influenced by spatial context by recovering precise cell-type localization.

Scientific Applications:

• Define spatial components of cellular identity: Maps cell-type composition to define spatial components of cellular identity within tissues.
• Uncover principles of cellular organization: Elucidates principles of cellular organization by resolving spatial distributions of cell types.
• Investigate spatially dependent gene expression: Investigates gene expression patterns that depend on spatial environment and cell-type localization.

Methodology:

Integrates cell-type profiles from single-cell RNA sequencing with spatial transcriptomic data to decompose mixed signals while accounting for technical variability across platforms.