spagg

Overview

The goal of spagg is to provide tools for aggregating spatial summary statistics generated from multiple regions-of-interest (ROIs) collected from the same tissue sample using multiplexed spatial proteomics technologies.

The analysis of spatial proteomics often involves calculating a spatial summary statistic, such as Ripley’s K, to quantify the level of clustering, repulsion, or complete spatial randomness exhibited by the cells. Given multiple ROIs for each sample, aggregating the spatial summary statistics for each ROI into a single value can facilitate downstream association testing with clinical outcomes.

spagg provides several methods for aggregating spatial summary statistics. These include three weighted means (diggle.avg, baddeley.avg, and landau.avg) and two ensemble approaches (ensemble.avg andcombo.weight.avg). The weighted means aggregate the summary statistics using a weighted mean based on the number of cells in each ROI and/or the area of each ROI. The ensemble approaches use random weights to construct an aggregation and combine the resulting p-values across many randomly-generated weights for an ensemble test.

The weighted means can be easily incorporated into other analytical approaches for spatial proteomics, such as the SPatial Omnibus Test (SPOT). Incorporation of the ensemble approaches in SPOT is in development.

Installation

You can install the development version of spagg from GitHub with:

# First, install devtools
if (!require("devtools", quietly = TRUE)) {
  install.packages("devtools")
}

# Install from Github
devtools::install_github("sarahsamorodnitsky/spagg")

spagg relies on several required dependencies: ACAT, dplyr, magrittr, the spatstat family of packages, stats, survival, and tidyselect. The ACAT package is currently in development on Github. To properly install spagg, the latest version of devtools is required to ensure it can install ACAT from Github.

In addition, spagg relies on several suggested packages to run the vignettes: knitr, rmarkdown, ggplot2, tidyr, spatstat.utils, spatstat.univar, cowplot, and SPOT. SPOT is also available on Github, so the latest version of devtools is required.

Vignettes

For example usage of spagg in single-cell spatial proteomics imaging analysis, please see the associated vignettes. The spagg package contains three vignettes:

1. Getting Started: this vignette illustrates how to use spagg for univariate or bivariate colocalization analyses on simulated data.
2. Analysis of a Non-Small Cell Lung Cancer Dataset: this vignette illustrates how to use spagg to analyze a non-small cell lung cancer dataset.
3. Accommodating Multiple Radii: this vignette illustrates how to incorporate multiple radii into a spagg analysis using the SPOT method.

Bugs and Improvements

Please feel free to use the Issues tab on the spagg Github site to note bugs or to suggest improvements.