SPOTlight

SPOTlight deconvolutes spatial transcriptomics spots by integrating spatial transcriptomics data with single-cell RNA sequencing data to infer the location of cell types and states within complex tissues.

Key Features:

• Seeded NMF Regression: Performs deconvolution using a seeded non-negative matrix factorization approach initialized with cell-type marker genes.
• NNLS Deconvolution: Applies non-negative least squares to subsequently deconvolute spatial transcriptomics capture locations.
• Visualization Tools: Provides visualization capabilities to assess deconvolution results.

Scientific Applications:

• Spatial Transcriptomics Deconvolution: Resolves cell-type composition and states within spatially indexed datasets that lack single-cell resolution.
• Tissue Organization Analysis: Enables inference of cell type and state locations to understand tissue organization and function.

Methodology:

SPOTlight uses a seeded non-negative matrix factorization (NMF) regression initialized with cell-type marker genes, followed by non-negative least squares (NNLS) to deconvolute spatial transcriptomics capture locations.