Spatstat

Spatstat performs spatial statistical analysis of point pattern data, primarily in two dimensions with limited three-dimensional and basic space-time support, to quantify spatial distributions and interactions of biological entities such as cells.

Key Features:

• Spatial point pattern analysis: Analyzes arrangements and interactions of points (for example, single cells) in predominantly 2D spatial data.
• Statistical functions: Implements functions including Kcross and pcf.cross for cross-type K-function and pair correlation analyses to assess clustering and inhibition.
• Dimensional versatility: Supports primarily 2D analyses with limited 3D and basic space-time analytical capabilities.
• Clustering and interaction testing: Provides statistical tests to detect clustering and inhibitory interactions between distinct feature types.
• Compatibility with image-derived single-cell data: Operates on coordinates derived from single-cell segmentation of whole-slide imaging, including multiplexed immunofluorescence data.

Scientific Applications:

• Tumor immune microenvironment analysis: Quantifies spatial distribution of cytotoxic T lymphocytes (CTL) relative to hypoxic and normoxic tumor regions, as applied to head and neck tumor specimens.
• Multiplex immunofluorescence studies: Analyzes whole-slide 4-plex immunofluorescence data to evaluate single-cell spatial patterns and interactions.
• Spatial hypothesis testing: Objectively assesses clustering and inhibitory interactions between cell types and microenvironmental features to support biological inference.

Methodology:

Whole-slide scans stained with 4-plex immunofluorescence were analyzed after single-cell segmentation using QuPath; specimens were examined for clustering and interactions between CTL, normoxic, and hypoxic tumor areas using Spatstat statistical tools including Kcross and pcf.cross to assess inhibitory interactions.