tests/testthat/test-merge_two_expfiles.R

In NathanSkene/EWCE: Expression Weighted Celltype Enrichment

# test merge_two_expfiles
test_that("merging two expression files", {
    # use vignette data
    if (!is_32bit()) {
        cortex_mrna <- ewceData::cortex_mrna()
        # NOTE: Can't use hypothalamus data for test as issue
        # with travis and read_xl::read_excel()

        # Download the hypothalamus data and unzip
        # if(!file.exists("GSE74672_expressed_mols_with_classes.xlsx")){
        #  download.file("ftp://ftp.ncbi.nlm.nih.gov/geo/series/GSE74nnn/GSE74672/suppl/GSE74672_expressed_mols_with_classes.xlsx.gz", destfile="GSE74672_expressed_mols_with_classes.xlsx.gz")
        #  system("gunzip GSE74672_expressed_mols_with_classes.xlsx.gz")
        # }

        # Read in the hypothalamus data
        # library(readxl)
        # hypo_dat = read_excel("GSE74672_expressed_mols_with_classes.xlsx")

        # Extract the expression data, gene symbols and annotation data
        # exp = data.matrix(hypo_dat[12:dim(hypo_dat)[1],2:dim(hypo_dat)[2]])
        # rownames(exp) = data.frame(hypo_dat[12:dim(hypo_dat)[1],1])[,1]
        # level1class = data.frame(level1class=t(hypo_dat[1,2:dim(hypo_dat)[2]]),stringsAsFactors = FALSE)[,1]
        # level2class = data.frame(leve2class=t(hypo_dat[2,2:dim(hypo_dat)[2]]),stringsAsFactors = FALSE)[,1]
        # cell_id     = colnames(hypo_dat)[2:dim(hypo_dat)[2]]
        # hypo_annot  = data.frame(cell_id=cell_id,level1class=level1class,level2class=level2class,stringsAsFactors = FALSE)

        # Drop the glia and unclassified cells (which don't have level 2  annotations)
        # hypo_annot  = hypo_annot[!is.na(hypo_annot$level2class) & !hypo_annot$level2class=="uc",]
        # hypo_exp    = exp[,hypo_annot$cell_id]

        # Make the cell type names more aesthetically pleasing
        # hypo_annot$level2class=gsub(",",";",hypo_annot$level2class)
        #  hypo_annot$level1class[grep("Oxt;|^Avp",hypo_annot$level2class)] = "Oxytocin / Vasopressin Expressing Neurons"
        #  hypo_annot$level1class[grep("^Th;|^Dopamine",hypo_annot$level2class)] = "Hypothalamic Dopaminergic Neurons"
        #  hypo_annot$level1class[grepl("^Vglut2|^Trh|^Qrfp|^Hcrt|^Pmch|^Adcyap1|^Npvf|^Ghrh|^Hmit|^Nms|^Vip;|^Per2|Tnr$|^Gad-low;Gnrh",hypo_annot$level2class) & grepl("neurons",hypo_annot$level1class)] = "Hypothalamic Glutamatergic Neurons"
        #  hypo_annot$level1class[grepl("GABA|^Sst|^Crh|^Npy|^Pomc|^Galanin|^Otof|Pnoc$|^Calcr-high",hypo_annot$level2class) & grepl("^neurons$",hypo_annot$level1class)] = "Hypothalamic GABAergic Neurons"
        #  hypo_annot$level2class[hypo_annot$level2class!=""] = sprintf("Hypothalamic %s Neuron",hypo_annot$level2class[hypo_annot$level2class!=""])

        # Fix bad MGI symbols
        # hypo_exp_CORRECTED = fix_bad_mgi_symbols(hypo_exp)
        # Merge the datasets
        merged_KI <- EWCE::merge_two_expfiles(
            exp1 = cortex_mrna$exp, # hypo_exp,#hypo_exp_CORRECTED,
            exp2 = cortex_mrna$exp,
            annot1 = cortex_mrna$annot, # hypo_annot,
            annot2 = cortex_mrna$annot,
            name1 = "Hypothalamus (KI)",
            name2 = "Cortex/Hippo (KI)"
        )

        # check merge didn't drop any samples
        test1 <- all.equal(
            sort(colnames(merged_KI$exp)),
            sort(c(
                colnames(cortex_mrna$exp),
                colnames(cortex_mrna$exp)
            ))
        )
        # remove folder once tested
        # unlink("GSE74672_expressed_mols_with_classes.xlsx")

        # fail if either function didn't create directory and add files
        testthat::expect_equal(test1, TRUE)
    }
})