B1096
Title: Bioconductor infrastructure for analyzing multiplex single cell imaging data in R
Authors: Julia Wrobel - Colorado School of Public Health (United States) [presenting]
Simon Vandekar - Vanderbilt University (United States)
Coleman Harris - Vanderbilt University Medical Center (United States)
Abstract: A new software ecosystem in R is introduced for data storage and spatial analysis of multiplex single-cell tissue imaging data. Our central R package is called spatialMI and builds on SpatialExperiment, which is an R package and S4 class on Bioconductor that provides a special data infrastructure for spatially resolved transcriptomics data that facilitates data storage, retrieval, subsetting, and interfacing with downstream tools. The spatialMI package adapts data structures from the SpatialExperiment class and inherits methods from that and the popular SingleCellExperiment class so that spatialMI users can easily access software developed for other similar single cell data types. For multiplex single-cell imaging data specifically, we build additional S4 methods to convert multichannel tiff images to a novel spatialMI data class so data from any multiplex platform, including CODEX, Vectra-Polaris, MIBI, IMC, etc, can be converted into the same type of Bioconductor data object. This facilitates consistency and ease of use in downstream analysis. We discuss the available functionality as well as forthcoming extensions in the form of satellite packages.
