Function reference • deconverse

Reference

Functions to build and interact with (hierarchical) single-cell reference objects.

new_screference(): Generate single-cell reference (`screference`) object

new_hscreference(): Generate multi-level hierarchical single-cell reference (`hscreference`) object

compute_reference(<screference>) compute_reference(<hscreference>): Compute reference given a deconvolution method from an `screference` object

autogenes_scref(): Compute reference matrix from an `screference` object using AutoGeneS

bayesprism_scref(): Compute reference matrix from an `screference` object using BayesPrism

card_scref(): Construct the mean gene expression basis matrix (B) from `screference` object using the method implemented in the CARD package

cell2location_scref(): Compute Reference Signature Matrix from `screference` Object Using Cell2location

cibersortx_scref(): Compute reference signature matrix from `screference` object using method implemented in CIBERSORTx

dwls_scref(): Compute reference signature matrix from `screference` object using method implemented in DWLS package for methods `dwls`, `svr` and `ols`

scaden_scref(): Compute reference matrix from an `screference` object using scaden (Single-cell assisted deconvolutional network)

spotlight_scref(): Select markers for SPOTlight using an `screference`

Benchmarking

Functions to create a benchmarking scbench object.

new_scbench(): Generate `scbench` object

mixtures_lod(): Simulate mixtures to find the limit of detection of each population in mixtures in an `scbench` object

mixtures_population(): Simulate sample population mixtures given bounds in an `scbench` object

mixtures_spatial(): Simulate spatially-aware mixtures in an `scbench` object

mixtures_spillover(): Simulate mixtures for measuring spillover between all pairs of populations in an `scbench` object

pseudobulks(): Generate pseudobulk gene expression from single-cell RNA-seq given population mixtures in an `scbench` object

Deconvolution

Functions to run deconvolution based on a single-cell reference

deconvolute(<scbench>) deconvolute(<matrix>) deconvolute(<Seurat>): Deconvolute `scbench` object using a chosen method

deconvolute_all(<scbench>) deconvolute_all(<matrix>) deconvolute_all(<Seurat>): Deconvolute `scbench` object using all methods with default settings

autogenes_deconvolute(): AutoGeneS deconvolution of bulk data using an `screference`

bayesprism_deconvolute(): BayesPrism deconvolution of bulk data using an `screference`

bisque_deconvolute(): Bisque deconvolution of bulk data using an `screference`

card_deconvolute(): CARD deconvolution using an `screference`

cell2location_deconvolute(): Cell2location deconvolution of spatial transcriptomics data from an 'screference'

cibersortx_deconvolute(): CIBERSORTx deconvolution of bulk data using an `screference`

dwls_deconvolute(): Dampened weighted least squares (DWLS) deconvolution of bulk data using an `screference`

music_deconvolute(): MuSiC deconvolution of bulk data using an `screference`

ols_deconvolute(): Ordinary least squares (OLS) deconvolution of bulk data using an `screference`

rctd_deconvolute(): RCTD deconvolution using an `screference`

scaden_deconvolute(): scaden (Single-cell assisted deconvolutional network) deconvolution of bulk data using an `screference`

spacet_deconvolute(): SpaCET deconvolution using an `screference`

spotlight_deconvolute(): SPOTlight deconvolution of spatial data using an `screference`

svr_deconvolute(): Support vector regression (SVR) deconvolution of bulk data using an `screference`

Plotting and Results

Functions to recover benchmarking results

results(<scbench>): Get summary of benchmark results for an analysis type

results_tidy(): Get benchmark results from `scbench` object in a tidy format, without performance metrics

plt_colocalization(): Plot spatial correlation of cell type proportions

plt_comp_performance(<scbench>) plt_comp_performance(<hscreference>) plt_comp_performance(<screference>): Plot computational metrics computed during deconvolution

plt_cor_heatmap(): Plot heatmap of correlation between deconvolution results for simulated populations and true mixtures used for the pseudobulks

plt_cors_scatter(): Plot correlations by population between deconvolution results for simulated populations and true mixtures used for the pseudobulk

plt_lod_heatmap(): Plot heatmap summarizing limit of detection by population and method

plt_lod_scatter(): Plot limit of detection of deconvolution for each population as the proportion of the pseudobulk where the method estimation is higher than at proportion=0 for each population

plt_method_concordance(): Plot percentage of concordance of major population between methods on spatial data

plt_method_correlation(): Plot correlation between methods on spatial data

plt_population_mixtures(): Plot sample population mixtures from an `scbench` object

plt_rmse_heatmap(): Plot RMSE between deconvolution results for simulated populations and true mixtures used for the pseudobulks

plt_spatial_correlation(): Plot spatial correlation of cell type proportions

plt_spatial_mixture(): Plot sample spatial mixture from an `scbench` object

plt_spillover_heatmap(): Plot heatmap summarizing spillover RMSE by population and method

plt_spillover_scatter(): Plot correlations between deconvolution results and spillover mixtures between each pair of populations

deconverse_results(<Seurat>) deconverse_results(<scbench>): Returns feature names of `deconverse` results that are stored in the Seurat object

Extras

Other package functions

pp_PBMCs(): Get Seurat PBMCs example and pre-process

read_h5ad(): Read a h5ad file and convert it to a Seurat object

test_markers(): Evaluate annotation labels using Seurat markers

install_cibersortx(): Installs CIBERSORTx container using docker or singularity

adata_to_seurat(): Convert adata to a Seurat object

seurat_to_adata(): Convert Seurat to a adata