Cardinal Scripts

A small collection of scripts written to streamline data processing/analysis in Cardinal.

Installation

Install from Github

devtools::install_github("gtluu/cardinalscripts")

Load from Local Folder

devtools::load_all("/path/to/cardinalscripts")

Dependencies

• R (>= 4.3.1)
• Cardinal (>= 3.4.1)
• stringr(>= 1.5.0)
• fuzzyjoin (>= 0.1.6)
• dplyr (>= 1.1.3)
• ggplot2 (>= 3.4.4)

Usage

updateMetadata

Use this function to add region of interest (ROI) and/or condition/treatment information based on x-y coordinates. An example Spot List (exported from Bruker flexImaging) and ROI table can be found in the examples folder.

data <- readMSIData('data.imzML')
data <- updateMetadata(data, spotsFile="spots.txt", roiFile="rois.csv")

subsetData

Use this function to remove any unwanted ROIs from your dataset. This will remove any pixel from that ROI from the MSImagingExperiment object.

data <- readMSIData('data.imzML')  # dataset with 4 ROIs
roi1 <- subset(data, c('roi2', 'roi3', 'roi4'))
roi2 <- subset(data, c('roi1', 'roi3', 'roi4'))
roi3 <- subset(data, c('roi1', 'roi2', 'roi4'))
roi4 <- subset(data, c('roi1', 'roi2', 'roi3'))

ionImageReport

Use this function to generate a PDF containing ion images for each feature from a vector of features.

ionImageReport(data, mz=mz(data), filename='ion_images')

sscImageReport

Use this function to generate a PDF containing segmentation maps for each model that was used for multivariate segmentation via spatial shrunken centroids (i.e. each combination of r, k, and s parameters).

sscImageReport(ssc, filename='segmentation_maps')

optimizeSSCParams

Use this function to assist in optimize the spatial neighborhood radius (r), initial number of segments (k), and sparsity (s) parameters used for running spatial shrunken centroids segmentation.

ssc <- spatialShrunkenCentroids(preprocessed_data, r=c(1, 2), k=c(5, 10, 20), s=c(0, 3, 6, 9))
ssc_params <- optimizeSSCParams(ssc)

getStatisticTable

Use this function to get a data.frame of values for a given set of r, k, and s parameters that can be used to tell which features contribute significantly to specific classes in a given spatial shrunken centroids segmentation model.

stat_table <- getStatisticTable(ssc, ssc_params$params)