stromal_subclasses

code for data processing and visualisation associated with the Wu et al. (2020) study "Stromal cell diversity associated with immune evasion in human triple‐negative breast cancer"

• https://www.embopress.org/doi/full/10.15252/embj.2019104063

Data availability

Processed data, including count matrices (raw and normalised) and cell metadata, can be found at the Broad Single Cell Portal at the following link:

• https://singlecell.broadinstitute.org/single_cell/study/SCP1106/stromal-cell-diversity-associated-with-immune-evasion-in-human-triple-negative-breast-cancer

Code Summary

01 cellranger count processing

job submission script for single-cell RNA-Seq processing using cellranger v2.1.1.

02 seurat v2 processing of individual datasets

job submission script and R script for processing individual seurat objects

03 seurat v2 data integration

job submission script and R script for integrating seurat objects

04 cluster annotation and reclustering

job submission script and R script for cluster annotations and reclustering of epithelial cells, stromal-immune cells and individual stromal subpopulations (CAFs and PVL cells). This R script also includes the generation of stromal gene signatures and export of gene expression matrices for downstream SCENIC (step 06).

05 gene signature analysis

AUCell

job submission script and R scripts for scoring of stromal-immune cells with cell type signatures (XCell database), T-cell exhaustion signatures (Blackburn et al. 2008) and the pancreatic ductal adenocarcinoma CAF signatures (David Tuveson's lab) using the AUCell method

clusterProfiler

job submission script and R scripts for gene ontology enrichment of the gene signatures for each stromal subcluster. Top 250 DEGs are used.

06 pySCENIC transcription factor enrichment of reclustered CAFs and PVL cells

job submission script and R scripts for exporting, filtering gene expression matrices, and running pySCENIC (python based command line version) with the CAFs and PVL raw count expression matrix as input. This also includes R script for filtering top TF candidates for clustering and visualisation

07 stromal cell signalling predictions

R script visualisation scripts for filtering stromal cell-cell signalling predictions.

Other analytical tools used

Immune evasion using TIDE

For computing T-cell dysfunction and exclusion analysis, please see the tumour immune dysfunction and exclusion (TIDE) method

• paper: https://www.nature.com/articles/s41591-018-0136-1
• link to online tool: http://tide.dfci.harvard.edu/

Cell signalling using NATMI

For processing ligand-receptor analysis, please see the NATMI documentation at:

• https://github.com/asrhou/NATMI

Contact

Please email [email protected] or [email protected] for any additional questions about the analytical methods used in this paper. All other relevant data are available from the authors upon request.