2023-cytomarkerpaper-analysis

Code to reproduce the figures and results from the cytomarker paper.

Data directory configuration

Prior to running, the following data files need to be placed in the data directory:

• singlecells.csv from the cytomarker paper data (download from zenodo here: https://zenodo.org/records/13891857)
• The nygc multimodal pbmc data set from: https://datasets.cellxgene.cziscience.com/de42a173-458a-429c-b129-c26bcd3adb3b.h5ad, named as nygc-pbmc.h5ad
• The transcriptome and proteome data from Nicolet et al. 2022 from https://doi.org/10.1371/journal.pone.0276294.s006
• The protein-RNA correlation table from the Gygi Lab here: https://gygi.hms.harvard.edu/data/ccle/Table_S4_Protein_RNA_Correlation_and_Enrichments.xlsx, put into the depmap sub-directory in data

Running

To run all:

cd 2023-cytomarkerpaper-analysis
snakemake --cores all # user can specify number of cores

Outputs:

1. data/sce_screen_full.rds and data/sce_screen_subsample.rds
2. /figs/screen_cluster_heatmap_unscaled.png and /figs/screen_cluster_heatmap_scaled.png to be used in cluster interpretation. User should then create data/cluster-interpretation-nov23.xlsx (what clusters are what cell types)
3. results/nygc_pbmc_subsampled.rds
4. figs/screen-vs-scrna.pdf
5. figs/rna-protein-scatter.pdf and figs/cytomarker_sens_spec_no_ms.pdf
6. figs/heatmap_mammary_single_cell.pdf