Balagan is an R-package dedicated to the statistical analysis of Multiplexed Imaging (MI) datasets. It contains several tools allowing to :
- Normalize and process the original data, i.e transforming and clustering the data.
- Infer the best parameters for an optimal spatial sampling strategy.
- Perform statistically powerful differential abundance analysis.
- Simulate stratified sampling using large multiplexed imaging panorama.
Balagan is based on the SingleCellExperiment object structure and is therefore compatible with a variety of other single-cell analysis tools. This package is aimed to provide advanced and robust statistical tool for the analysis of MI and therefore I strongly recommend the user to read all the mathematical papers mentionned in the documentation.
Scripts initially written for our Nature Method paper have been integrated to this package.
Balagan can be installed from the source file :
devtools::install_local("Path/to/balagan_1.0.0.tar.gz",dependencies = T)It can also be installed using devtools :
devtools::install_github("PierreBSC/Balagan")Before performing any analysis using Balagan, we strongly recommend you to be familiar with some key statistical concepts. Below is a list of papers, notes and book chapters that :
- Our paper on random sampling.
- Our new biorxiv preprint on the statistical modeling of cell count data obtained by multiplexed imaging.
- The mathematical notes associated with our new preprint.
- For an introduction to Generalized Linear Models, chapter 4 of "An Introduction to Statistical Learning".
- The excellent textbook on spatial point pattern "Statistical analysis and modelling of spatial point patterns". While reading the 557 pages of this book is not necesseray, key notions (comlplete spatial randomness, homogenous poisson point pattern...) are explained in a simple yet rigorous manner in this book
As mentionned above, Balagan can be used to perform various tasks and comprehensive tutorials have been written for each of them: