Methods for differential abundance analyses of mass cytometry data

To repeat the real data analyses for the MEF reprogramming time course:

• Download the FCS files from Cytobank for accession number 43324, and put them in refdata/Cytobank_43324.
• Enter refdata and run preprocess_43324.R to transform and gate on the intensities, followed by gen_objects_43324.R to obtain the hypersphere counts.
• Enter real/analysis and run (in order) detect_da.R, to detect DA hyperspheres; plot_da_initial.R, to construct the t-SNE plots; plot_da_extra.R, to construct subpopulation- and time-point-specific plots; and plot_dispersions.R, to construct the dispersion plot. Note that the t-SNE plots are stochastic and may not be the same as in the paper.

Alternatively, you can look at workflow.Rmd in real/workflow, which provides an annotated workflow of the processing and analysis of the Oct4-GFP time course data.

To run the additional analyses for the time course data:

• Enter real/neighbours and run get_neighbors.R to construct plots of distances to nearest neighbours.
• Enter real/robustness and run robustness.R to obtain DA statistics with different parameter settings.
• Enter real/clustering and run citrus_test.R to obtain DA clusters from CITRUS. Then run map_onto_tsne_citrus.R to map the CITRUS cluster centres onto the t-SNE plot from the hypersphere analysis.

To run the analyses for the BMMC data:

• Download the FCS files from Cytobank for accession number 44185, and put them in refdata/Cytobank_44185.
• Enter refdata and run preprocess_44185.R to transform and gate on the intensities, followed by gen_objects_44185.R to obtain the hypersphere counts.
• Enter real/secondary and run (in order) detect_da.R, to detect DA hyperspheres; and plot_da.R, to construct the t-SNE plots.

To repeat the simulations, enter simulations and run:

• edgeR_check.R, to check type I error control.
• FDR_check.R, to check spatial FDR control.
• cluster_sim.R, to check performance of clustering.
• radius_sim.R, to check the effect of increasing the radius on hypersphere positions.
• shift_check.R, to check the type I error control on intensity-shifted data.