Merge pull request #25 from BodenmillerGroup/devel

More infos and how to write out compensated images

Author: nilseling

.github/workflows/build.yml

@@ -32,7 +32,7 @@ jobs:
                                             "distill", "openxlsx", "ggrepel", "patchwork", "mclust",
                                             "RColorBrewer", "uwot", "Rtsne", "harmony", "Seurat", "SeuratObject", "cowplot",
                                             "kohonen", "caret", "randomForest", "ggridges", "cowplot", "gridGraphics", 
-                                            "scales"))'
+                                            "scales", "tiff"))'
       - name: Install Bioconductor dependencies
         run: Rscript -e 'BiocManager::install(c("CATALYST", "scuttle", "scater", "dittoSeq", "tidyverse", "BiocStyle", "batchelor", "bluster",
                                                 "scran", "lisaClust", "spicyR"))'

05-spillover_matrix.Rmd

@@ -369,6 +369,21 @@ For convenience, we will re-set the `channelNames` to their biological targtes:
 channelNames(images_comp) <- rownames(spe)
 ```
 
+## Write out compensated images
+
+In the final step, the compensated images are written out as 16-bit TIFF
+files:
+
+```{r write-images, message=FALSE, results='hide'}
+library(tiff)
+dir.create("data/comp_img")
+lapply(names(images_comp), function(x){
+  writeImage(as.array(images_comp[[x]])/(2^16 - 1), 
+             paste0("data/comp_img/", x, ".tiff"),
+             bits.per.sample = 16)
+})
+```
+
 ## Save objects
 
 For further downstream analysis, the compensated `SpatialExperiment` and

README.md

@@ -8,22 +8,22 @@ R workflow highlighting analyses approaches for imaging mass cytometry (IMC; or
 
 ## Scope
 
-**This project is under development and will change on a regular basis**
 
-Upon release, this workflow highlights the use of common R/Bioconductor packages to pre-process and analyse single-cell data obtained from segmented IMC images.
+This workflow explains the use of common R/Bioconductor packages to pre-process and analyse single-cell data obtained from segmented IMC images.
 While we use IMC data as an example, the concepts presented here can be applied to images obtained by other technologies (e.g. CODEX, MIBI, mIF, etc.).
 The workflow can be largely divided into the following parts:
 
-1. Preprocessing (reading in the data, spillover correctio)
-2. Metrics for quality control
-3. Low-dimensional visualization, clustering and/or cell-type classification
-4. Visualization of cell- and pixel-level information
-5. Spatial analyses
+1. Preprocessing (reading in the data, spillover correction)
+2. Image- and cell-level quality control, low-dimensional visualization
+3. Sample/batch effect correction
+4. Cell phenotyping via clustering or classification
+5. Image visualization
+6. Spatial analyses
 
 ## Usability
 
 After cloning the repository, the code can be run as is.
-It is continously tested on a ubuntu system using the newest release versions of the used packages.
+It is continously tested on a ubuntu system using the newest release versions of the used R packages.
 
 ## Contribution
 
@@ -32,4 +32,23 @@ You can also add sections by forking the repo, adding your changes and issuing a
 
 ## Maintainer
 
-[Nils Eling](https://github.com/nilseling)
+[Nils Eling](https://github.com/nilseling)
+
+## Contributors
+
+[Vito Zanotelli](https://github.com/votti)
+[Daniel Schulz](https://github.com/SchulzDan)
+[Jonas Windhager](https://github.com/jwindhager) 
+[Michelle Daniel](https://github.com/michdaniel)
+
+## Citation
+
+The workflow is currently under development for final publication.
+In the meantime please refer to the 
+[preprint](https://www.biorxiv.org/content/10.1101/2021.11.12.468357v1) 
+which you can site as follows:
+
+```
+Jonas Windhager, Bernd Bodenmiller, Nils Eling (2020). An end-to-end workflow for multiplexed image processing and analysis. 
+    bioRxiv, doi: 10.1101/2021.11.12.468357
+```

index.Rmd

@@ -13,9 +13,7 @@ description: "This bookdown project highlights possible down-stream analyses per
 
 # IMC Data Analysis Workflow {#preamble}
 
-**This project is under development and will change on a regular basis**
-
-Upon release, this workflow highlights the use of common R/Bioconductor packages
+This workflow highlights the use of common R/Bioconductor packages
 to analyze single-cell data obtained from segmented imaging mass cytometry (IMC)
 images. We will not perform multi-channel image processing and segmentation in R
 but rather link to available approaches in Section \@ref(processing). While we
@@ -30,11 +28,12 @@ strategies to extract single-cell data from IMC images in Section
 Reproducible code written in R is available from Section \@ref(prerequisites)
 onwards and the workflow can be largely divided into the following parts:
 
-1. Pre-processing (reading in the data, spillover correction)
-2. Metrics for quality control
-3. Low-dimensional visualization, clustering and/or cell-type classification
-4. Visualization of cell- and pixel-level information
-5. Spatial analyses
+1. Preprocessing (reading in the data, spillover correction)
+2. Image- and cell-level quality control, low-dimensional visualization
+3. Sample/batch effect correction
+4. Cell phenotyping via clustering or classification
+5. Image visualization
+6. Spatial analyses
 
 ## Feedback and contributing