Script descriptions

Input files are in folder -- input folder (Seurat object needs to be downloaded from Zenodo) Output files are default to -- output folder Utility functions used in the codes are in folder -- util_fun folder pesudobulk scripts -- pbulk_DE folder BCR scripts -- BCR folder TCR scripts -- TCR folder

1.demux.preprocess.R

This script takes the processed data from cellranger and demuxlet files. Creats a Seurat object, performs QC filtering of the cells, hashtag demultiplexing and DSB normalization of the CITEseq data.

2.clustering.R

This script takes the Seurat object output by 1.demux.preprocess.R, and performs clustering using the cell surface antibody data (ADT/CITE) and adds the cell annotations.

3.1.label.transfer.from.3batches.R

Generates Extended Data Fig.5a
This script uses the data from adult COVID data (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE161918) and integrates and predicts celltype labels of this dataset from annotated adult COVID dataset.

3.2.cell_no_check_all4batches.R

Generates Extended Data Fig.5b
Cell frequency summary taking the predicted labels and adult COVID dataset to compare interested cell populations in both adult and pediatric patients

4.meta.plot.R

Generates Fig.4a and Extended Data Fig.5d
This script takes the Seurat object (uploaded) and plots the UMAP visulization of total cells and Monocytes. Also generates single cell CD163 expression of monocytes and S100A family genes expression plots.

pesudobulk scripts are in pbulk_DE folder

Main workflow for the pseudobulk differential expresssion analyses are as follows:

1.pbulk.pooling.unfiltered.R

Pool reads from single cells together per celltype per sample

2.calcnormfactors_and_filter_genes_samples.R

Caculate normalization factors, normalize pseudobulk libraries and filter genes and samples

3.make_design_matrix.R

make design matrix for differential expression analyses

4.run_de_limma.R

Run limma differential expression model

5.make_contrast_fit.R

Make contrast matrix for each celltype and get top table of differentially expressed genes

6.run_fgsea.R

Run fgsea for the pseudobulk differential expresssion results using the t-statistic ranks

7.plot_fgsea.R

Generates Fig.4b, 4c, Extended Data Fig.5c
Fgsea results visulization of selected pathways

8.DEplot.subject.hm.R

Generates Fig.4d
Heatmaps of HALLMARK_TNFa_Signaling_via_NFkB gene set leading edge genes. Showing CD4 Memory and Classical Monocytes as examples.

9.gsva.score.miscvscovid.R

Generates Fig.4e
Calculate GSVA scores using leading edge genes of MIS-C vs pCOVID-19. Take input data from normalized pseudobulk objects.
Use GSVA scores to genereate HALLMARK_TNFa_Signaling_via_NFkB gene set signature score boxplot.

BCR scripts are in BCR folder

1.add.BCR.meta

Add BCR metadata to Seurat object.

2.BCRstat.wo.geneexpr.R

Generates Fig.5d, Extended Data Fig.5b, 5d
BCR IgH usage and B cell mutation frequency summary.

3.BCR.mut.corr.R

Generates Extended Data Fig.5e, 5f
Memory B cell and Plasmablast mutation frequencies correlation with cell surface protein.

TCR scripts are in TCR folder

1.compile.TCR.data.R

Compile raw TCR data and add metadata to the cells.

2.TCRstat.wo.geneexpr.R

Generates Fig.5c
CD4 TCR TRBV11-2 usage summary.

3.add.TCR.meta.R

Add TCR metadata to Seurat object.

4.TRBV112CD4.gex.R

Generates Extended Data Fig.6d, 6e
TRBV11-2 MIS-C CD4 T cell phenotype, compared to other TRBV11-2 negative CD4 T cells within MIS-C patients. Heatmaps of genes and surface proteins markers of MIS-C TRBV11-2 CD4 T cell.

5.TRBV112CD4.gsea.R

Generates Extended Data Fig.6f
Fgsea test on the genes from single cell level differential expression analysis of TRBV11-2 CD4 T cells within MIS-C patients. Plot of apoptosis related gene sets enrichment.