SpatialRNA

SpatialRNA generates graphs and subgraphs from spatial RNA molecules in tissue samples, facilitating the application of Graph Neural Network models for analysis of single-molecule spatial transcriptomics data.

Key Features:

• Graph generation: Generates graphs and subgraphs from spatial RNA molecules in tissue samples.
• GNN model application: Enables application of Graph Neural Network models under the PyG framework.
• Spatial domain segmentation: Comprehensively segments tissue into spatial domains or niches, identifying groups of molecules and/or cells that collaboratively work together to produce complex phenotypes.
• Scalability: Handles the vast number of detected transcripts in image-based spatial transcriptomics datasets.

Scientific Applications:

• Spatial domain detection: Detects spatial domains or niches in image-based spatial transcriptomics data at single-transcript and single-cell levels.
• Molecular microenvironment analysis: Aids biological interpretation of image-based spatial transcriptomics data and its underlying molecular microenvironments.
• Complex phenotype capture: Captures complex molecular and cellular phenotypes in tissues using Graph Neural Network models.

Methodology:

SpatialRNA constructs RNA molecule graphs from image-based spatial transcriptomics data and applies Graph Neural Network models under the PyG framework for downstream analysis.