tests/testthat/test_scEpigenomics.R

In vallotlab/IDclust: Iterative Differential Clustering for single-cell

context("Testing IDclust scEpigenomics")

# Functions for testing purposes
# if(requireNamespace("ChromSCape")){
#     set.seed(47)
#     out = create_scDataset_raw(featureType = "window",sparse = TRUE, 
#                                batch_id = factor(c(1,1,2,2)))
#     mat = out$mat
#     annot = out$annot
#     batches = out$batches
#     
#     # Download, extract & format PairedTag dataset - H3K27ac (Zhu et al., 2021)
#     temp = tempfile()
#     tempdir_1 = tempdir()
#     download.file("https://www.ncbi.nlm.nih.gov/geo/download/?acc=GSE152020&format=file&file=GSE152020%5FPaired%2DTag%5FH3K27ac%5FDNA%5Ffiltered%5Fmatrix%2Etar%2Egz", temp)
#     untar(temp, exdir = tempdir_1)
#     features = read.table(file.path(tempdir_1, "04.Paired-Tag_H3K27ac_DNA_filtered_matrix", "bins.tsv"),
#                           row.names = NULL, header = F, sep = "\t")[,1, drop = F]
#     write.table(features, file = file.path(tempdir_1, "04.Paired-Tag_H3K27ac_DNA_filtered_matrix", "features.tsv"),
#                 row.names = F, col.names = F, quote = F)
#     # Reading the matrix with ChromSCape
#     out = ChromSCape::read_sparse_matrix(file.path(tempdir_1, "04.Paired-Tag_H3K27ac_DNA_filtered_matrix"),
#                                          ref = "mm10", verbose = TRUE)
#     unlink(file.path(tempdir_1, "04.Paired-Tag_H3K27ac_DNA_filtered_matrix"), recursive = TRUE)
#     
#     scExp = ChromSCape::preprocessing_filtering_and_reduction(
#         datamatrix = out$datamatrix,
#         annot_raw = out$annot_raw,
#         min_reads_per_cell = 200,
#         max_quantile_read_per_cell = 99,
#         n_top_features = nrow(out$datamatrix),
#         norm_type = "TFIDF",
#         remove_PC = "Component_1",
#         subsample_n = NULL,
#         ref_genome = "mm10",
#         exclude_regions = NULL,
#         doBatchCorr = FALSE,
#         batch_sels = NULL
#         )
#     
#     scExp = find_clusters_louvain_scExp(scExp, k = 100, resolution = 0.1, use.dimred = "PCA")
#     
#     outdir = tempdir()
#     
#     scExp_IDC = iterative_differential_clustering_scEpigenomics(scExp, output_dir = outdir, nPCA = 10, runFDR = F)
#     
#     #test sparse matrix
#     test_that("Sparse matrices", {
#         scExp = create_scExp(mat,annot)
#         expect_is(SingleCellExperiment::counts(scExp),"dgCMatrix")
#         scExp = filter_scExp(scExp)
#         expect_is(SingleCellExperiment::counts(scExp),"dgCMatrix")
#         scExp = normalize_scExp(scExp,type = "CPM")
#         expect_is(SingleCellExperiment::normcounts(scExp),"dgCMatrix")
#         scExp=feature_annotation_scExp(scExp)
#         expect_is(SummarizedExperiment::rowRanges(scExp),"GRanges")
#         scExp = reduce_dims_scExp(scExp,n = 50,batch_correction = FALSE)
#         expect_is(SingleCellExperiment::reducedDim(scExp,"PCA"),"data.frame")
#         
#         scExp = colors_scExp(scExp,annotCol = c("sample_id","batch_id","total_counts"))
#         plot_reduced_dim_scExp(scExp,reduced_dim = "PCA",color_by = "sample_id")
#         
#         scExp = correlation_and_hierarchical_clust_scExp(scExp)
#         expect_is(SingleCellExperiment::normcounts(scExp),"dgCMatrix")
#         scExp = filter_correlated_cell_scExp(scExp)
#         expect_is(SingleCellExperiment::normcounts(scExp),"dgCMatrix")
#         scExp = consensus_clustering_scExp(scExp)
#         expect_is(SingleCellExperiment::normcounts(scExp),"dgCMatrix")
#         expect_is(scExp@metadata$consclust,"list")
#         expect_is(scExp@metadata$consclust[[2]]$consensusClass,"integer")
#         scExp = choose_cluster_scExp(scExp,nclust = 2)
#         expect_is(SingleCellExperiment::normcounts(scExp),"dgCMatrix")
#     })
# } else{
#     message("Testing IDclust scEpigenomic - no package ChromSCape. Skipping tests")
# }