4 Prepare scRNA-seq and ST data (Moncada et al., 2020)

4.1 scRNA-seq data

The data can be loaded by the following code:

pdacrna <- readRDS(url("https://figshare.com/ndownloader/files/34112468"))

The data are stored in DOI:10.6084/m9.figshare.19200254 and the generating process is described below.

From GSE111672, we downloaded inDrop data with sample accession numbers GSM3036909, GSM3036910, GSM3405527, GSM3405528, GSM3405529, and GSM3405530 (PDAC-A inDrop1-inDrop6).

fn <- c("rawdata/2020_001_Moncada/pdac_indrop/PDACA_1/results/gene_expression.tsv",
        "rawdata/2020_001_Moncada/pdac_indrop/PDACA_2/results/gene_expression.tsv",
        "rawdata/2020_001_Moncada/pdac_indrop/PDACA_3/results/gene_expression.tsv",
        "rawdata/2020_001_Moncada/pdac_indrop/PDACA_4/results/gene_expression.tsv",
        "rawdata/2020_001_Moncada/pdac_indrop/PDACA_5/results/gene_expression.tsv",
        "rawdata/2020_001_Moncada/pdac_indrop/PDACA_6/results/gene_expression.tsv")
pdacrna <- list()
for(i in seq_along(fn)){
  d <- read.table(fn[i], header = TRUE, stringsAsFactors = FALSE, row.names = 1)
  colnames(d) <- paste0("PDAC-A-inDrop", i, "-", colnames(d))
  pdacrna[[i]] <- SingleCellExperiment(assays = list(counts = as.matrix(d)),
                                       rowData = data.frame(gene = rownames(d)),
                                       colData = data.frame(sample = colnames(d)))
}

rbind(dim(pdacrna[[1]]), dim(pdacrna[[2]]), dim(pdacrna[[3]]),
      dim(pdacrna[[4]]), dim(pdacrna[[5]]), dim(pdacrna[[6]]))

      [,1]  [,2]
[1,] 19811 10000
[2,] 19811 10000
[3,] 19811 10000
[4,] 19811 10000
[5,] 19811 10000
[6,] 19811 10000

Add metadata for both variables and samples using ASURAT function add_metadata().

for(i in seq_along(pdacrna)){
  pdacrna[[i]] <- add_metadata(sce = pdacrna[[i]], mitochondria_symbol = "^MT-")
}

Qualities of sample (cell) data are confirmed based on proper visualization of colData(sce).

for(i in seq_along(pdacrna)){
  df <- data.frame(x = colData(pdacrna[[i]])$nReads,
                   y = colData(pdacrna[[i]])$nGenes)
  p <- ggplot2::ggplot() +
    ggplot2::geom_point(ggplot2::aes(x = df$x, y = df$y), size = 1, alpha = 1) +
    ggplot2::labs(title = paste0("PDAC-A inDrop ", i),
                  x = "Number of reads", y = "Number of genes") +
    ggplot2::theme_classic(base_size = 20) +
    ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 20)) +
    ggplot2::scale_x_log10(limits = c(1, 20000)) +
    ggplot2::scale_y_log10(limits = c(1, 10000))
  p <- ggExtra::ggMarginal(p, type = "histogram", margins = "both", size = 5,
                           col = "black", fill = "gray")
  filename <- paste0("figures/figure_08_0005_", i, ".png")
  ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 5)
}

Confirming that the data qualities are comparable among experimental batches, concatenate all the objects horizontally.

# Take intersection of genes.
genes <- Reduce(intersect, list(rownames(pdacrna[[1]]), rownames(pdacrna[[2]]),
                                rownames(pdacrna[[3]]), rownames(pdacrna[[4]]),
                                rownames(pdacrna[[5]]), rownames(pdacrna[[6]])))
for(i in seq_along(pdacrna)){
  pdacrna[[i]] <- pdacrna[[i]][genes, ]
  rowData(pdacrna[[i]])$nSamples <- NULL
}
# Horizontally concatenate SingleCellExperiment objects.
pdacrna <- cbind(pdacrna[[1]], pdacrna[[2]], pdacrna[[3]],
                 pdacrna[[4]], pdacrna[[5]], pdacrna[[6]])
# Add metadata again.
pdacrna <- add_metadata(sce = pdacrna, mitochondria_symbol = "^MT-")

dim(pdacrna)

[1] 19811 60000

4.2 ST data

The data can be loaded by the following code:

pdacst <- readRDS(url("https://figshare.com/ndownloader/files/34112471"))

The data are stored in DOI:10.6084/m9.figshare.19200254 and the generating process is described below.

Load a raw read count table, convert Ensembl IDs into gene symbols, and change the column names.

fn <- "rawdata/2020_001_Moncada/pdac_st/SRR6825057_stdata.tsv"
pdacst <- read.table(fn, header = TRUE, stringsAsFactors = FALSE, row.names = 1)
pdacst <- t(pdacst)
ensembl <- rownames(pdacst)
dictionary <- AnnotationDbi::select(org.Hs.eg.db::org.Hs.eg.db, key = ensembl,
                                    columns = c("SYMBOL", "ENTREZID"),
                                    keytype = "ENSEMBL")
dictionary <- dictionary[!duplicated(dictionary$ENSEMBL), ]
dictionary[which(is.na(dictionary$SYMBOL)),]$SYMBOL <- as.character("NA")
rownames(pdacst) <- make.unique(as.character(dictionary$SYMBOL))
colnames(pdacst) <- paste0("PDAC-A-ST1_", colnames(pdacst))

Create a SingleCellExperiment object by inputting the read count table.

pdacst <- SingleCellExperiment(assays = list(counts = as.matrix(pdacst)),
                               rowData = data.frame(gene = rownames(pdacst)),
                               colData = data.frame(sample = colnames(pdacst)))

A Seurat object, including PDAC tissue images, was obtained from the authors of DOI:10.1038/s41587-019-0392-8 and DOI:10.1093/nar/gkab043, and set the tissue image data into the metadata slot.

fn <- "rawdata/2020_001_Moncada/pdac_st/PDAC-A_ST_list.RDS"
pdacst_surt <- readRDS(file = fn)
pdacst_surt <- pdacst_surt$GSM3036911
metadata(pdacst)$images <- pdacst_surt@images

Since the above SingleCellExperiment object includes spatial coordinates outside of tissues, remove such spots.

pdacst <- pdacst[, colnames(pdacst_surt)]
identical(colnames(pdacst), colnames(pdacst_surt))

[1] TRUE

dim(pdacst)

[1] 25807   428

Add metadata for both variables and samples using ASURAT function add_metadata().

pdacst <- add_metadata(sce = pdacst, mitochondria_symbol = "^MT-")

Qualities of spot data are confirmed based on proper visualization of colData(sce).

df <- data.frame(x = colData(pdacst)$nReads, y = colData(pdacst)$nGenes)
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df$x, y = df$y), size = 1, alpha = 1) +
  ggplot2::labs(title = "PDAC-A ST", x = "Number of reads", y = "Number of genes") +
  ggplot2::theme_classic(base_size = 18) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 20)) +
  ggplot2::scale_x_log10(limits = c(-NA, NA)) +
  ggplot2::scale_y_log10(limits = c(-NA, NA))
p <- ggExtra::ggMarginal(p, type = "histogram", margins = "both", size = 5,
                         col = "black", fill = "gray")
filename <- "figures/figure_09_0005.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 5)

5 Preprocessing

5.1 Control data quality

Remove variables (genes) and samples (cells) with low quality, by processing the following three steps:

remove variables based on expression profiles across samples,
remove samples based on the numbers of reads and nonzero expressed variables,
remove variables based on the mean read counts across samples.

5.1.1 Remove variables based on expression profiles

ASURAT function remove_variables() removes variable (gene) data such that the numbers of non-zero expressing samples (cells) are less than min_nsamples.

pdacrna <- remove_variables(sce = pdacrna, min_nsamples = 10)
pdacst <- remove_variables(sce = pdacst, min_nsamples = 10)

5.1.2 Remove samples based on expression profiles

Qualities of sample (cell) data are confirmed based on proper visualization of colData(sce). ASURAT function plot_dataframe2D() shows scatter plots of two-dimensional data (see here for details).

title <- "PDAC-A inDrop"
df <- data.frame(x = colData(pdacrna)$nReads, y = colData(pdacrna)$nGenes)
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df$x, y = df$y), size = 1, alpha = 1) +
  ggplot2::labs(title = title, x = "Number of reads", y = "Number of genes") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 20))
filename <- "figures/figure_08_0010.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 5)

df <- data.frame(x = colData(pdacrna)$nReads, y = colData(pdacrna)$percMT)
title <- "PDAC-A inDrop"
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df$x, y = df$y), size = 1, alpha = 1) +
  ggplot2::labs(title = title, x = "Number of reads", y = "Perc of MT reads") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5))
filename <- "figures/figure_08_0011.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 5)

ASURAT function remove_samples() removes sample (cell) data by setting cutoff values for the metadata.

pdacrna <- remove_samples(sce = pdacrna, min_nReads = 1000, max_nReads = 10000,
                          min_nGenes = 100, max_nGenes = 1e+10,
                          min_percMT = 0, max_percMT = 20)

pdacst <- remove_samples(sce = pdacst, min_nReads = 0, max_nReads = 1e+10,
                         min_nGenes = 0, max_nGenes = 1e+10,
                         min_percMT = NULL, max_percMT = NULL)

5.1.3 Remove variables based on the mean read counts

Qualities of variable (gene) data are confirmed based on proper visualization of rowData(sce). ASURAT function plot_dataframe2D() shows scatter plots of two-dimensional data.

title <- "PDAC-A inDrop"
aveexp <- apply(as.matrix(assay(pdacrna, "counts")), 1, mean)
df <- data.frame(x = seq_len(nrow(rowData(pdacrna))),
                 y = sort(aveexp, decreasing = TRUE))
p <- ggplot2::ggplot() + ggplot2::scale_y_log10() +
  ggplot2::geom_point(ggplot2::aes(x = df$x, y = df$y), size = 1, alpha = 1) +
  ggplot2::labs(title = title, x = "Rank of genes", y = "Mean read counts") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5))
filename <- "figures/figure_08_0015.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 5)

title <- "PDAC-A ST"
aveexp <- apply(as.matrix(assay(pdacst, "counts")), 1, mean)
df <- data.frame(x = seq_len(nrow(rowData(pdacst))),
                 y = sort(aveexp, decreasing = TRUE))
p <- ggplot2::ggplot() + ggplot2::scale_y_log10() +
  ggplot2::geom_point(ggplot2::aes(x = df$x, y = df$y), size = 1, alpha = 1) +
  ggplot2::labs(title = title, x = "Rank of genes", y = "Mean read counts") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5))
filename <- "figures/figure_09_0015.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 5)

ASURAT function remove_variables_second() removes variable (gene) data such that the mean read counts across samples are less than min_meannReads.

pdacrna <- remove_variables_second(sce = pdacrna, min_meannReads = 0.01)
pdacst <- remove_variables_second(sce = pdacst, min_meannReads = 0.01)

rbind(dim(pdacrna), dim(pdacst))

[1,] 12248 2034
[2,] 10364  428

Check the number of genes, which commonly exist in both the datasets.

length(intersect(rownames(pdacrna), rownames(pdacst)))

[1] 7923

Qualities of spot data are confirmed based on proper visualization of colData(sce).

# scRNA-seq
df <- data.frame(x = colData(pdacrna)$nReads, y = colData(pdacrna)$nGenes)
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df$x, y = df$y), size = 1, alpha = 1) +
  ggplot2::labs(title = "PDAC-A inDrop",
                x = "Number of reads", y = "Number of genes") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 20)) +
  ggplot2::scale_x_log10(limits = c(1, 20000)) +
  ggplot2::scale_y_log10(limits = c(1, 10000))
p <- ggExtra::ggMarginal(p, type = "histogram", margins = "both", size = 5,
                         col = "black", fill = "gray")
filename <- "figures/figure_10_0005.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 5)
# ST
df <- data.frame(x = colData(pdacst)$nReads, y = colData(pdacst)$nGenes)
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df$x, y = df$y), size = 1, alpha = 1) +
  ggplot2::labs(title = "PDAC-A ST", x = "Number of reads", y = "Number of genes") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 20)) +
  ggplot2::scale_x_log10(limits = c(1, 40000)) +
  ggplot2::scale_y_log10(limits = c(1, 10000))
p <- ggExtra::ggMarginal(p, type = "histogram", margins = "both", size = 5,
                         col = "black", fill = "gray")
filename <- "figures/figure_10_0006.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 5)

5.2 Normalize data

Perform bayNorm() (Tang et al., Bioinformatics, 2020) for attenuating technical biases with respect to zero inflation and variation of capture efficiencies between samples (cells).

# pdacrna
bayout <- bayNorm::bayNorm(Data = assay(pdacrna, "counts"), mode_version = TRUE)
assay(pdacrna, "baynorm") <- bayout$Bay_out
# pdacst
bayout <- bayNorm::bayNorm(Data = assay(pdacst, "counts"), mode_version = TRUE)
assay(pdacst, "baynorm") <- bayout$Bay_out

Normalize the data using canonical correlation analysis-based method using Seurat functions (Butler Nat. Biotechnol., 2018).

surt <- list()
surt[[1]] <- Seurat::as.Seurat(pdacrna, counts = "baynorm", data = "baynorm")
surt[[2]] <- Seurat::as.Seurat(pdacst, counts = "baynorm", data = "baynorm")
names(surt) <- c("inDrop", "ST")
for(i in seq_along(surt)){
  surt[[i]] <- Seurat::NormalizeData(surt[[i]])
  surt[[i]] <- Seurat::FindVariableFeatures(surt[[i]], selection.method = "vst",
                                            nfeatures = 7500)
}
genes <- Seurat::SelectIntegrationFeatures(object.list = surt, nfeatures = 7500)
anchors <- Seurat::FindIntegrationAnchors(object.list = surt,
                                          anchor.features = genes)
surt <- Seurat::IntegrateData(anchorset = anchors,
                              normalization.method = "LogNormalize")
Seurat::DefaultAssay(surt) <- "integrated"
pdac <- Seurat::as.SingleCellExperiment(surt)
rowData(pdac) <- rownames(pdac)

Keep the tissue coordinate information.

pdac@metadata <- pdacst@metadata

dim(pdac)

[1] 6761 2462

Center row data.

mat <- assay(pdac, "logcounts")
assay(pdac, "centered") <- sweep(mat, 1, apply(mat, 1, mean), FUN = "-")

Set gene expression data into altExp(sce).

altExps(pdac) <- NULL  # For safely using Seurat function as.Seurat() later,
                       # avoid using the same slot names in assayNames and altExpNames.
sname <- "logcounts"
altExp(pdac, sname) <- SummarizedExperiment(list(counts = assay(pdac, sname)))

Add ENTREZ Gene IDs to rowData(sce).

dictionary <- AnnotationDbi::select(org.Hs.eg.db::org.Hs.eg.db,
                                    key = rownames(pdac),
                                    columns = "ENTREZID", keytype = "SYMBOL")
dictionary <- dictionary[!duplicated(dictionary$SYMBOL), ]
rowData(pdac)$geneID <- dictionary$ENTREZID

6 Multifaceted sign analysis

Infer cell or disease types, biological functions, and signaling pathway activity at the single-cell level by inputting related databases.

ASURAT transforms centered read count tables to functional feature matrices, termed sign-by-sample matrices (SSMs). Using SSMs, perform unsupervised clustering of samples (cells).

6.1 Compute correlation matrices

Prepare correlation matrices of gene expressions.

mat <- t(as.matrix(assay(pdac, "centered")))
cormat <- cor(mat, method = "spearman")

6.2 Load databases

Load databases.

urlpath <- "https://github.com/keita-iida/ASURATDB/blob/main/genes2bioterm/"
load(url(paste0(urlpath, "20201213_human_DO.rda?raw=TRUE")))         # DO
load(url(paste0(urlpath, "20201213_human_CO.rda?raw=TRUE")))         # CO
load(url(paste0(urlpath, "20220308_human_MSigDB.rda?raw=TRUE")))     # MSigDB
load(url(paste0(urlpath, "20220308_human_CellMarker.rda?raw=TRUE"))) # CellMarker
load(url(paste0(urlpath, "20201213_human_GO_red.rda?raw=TRUE")))     # GO
load(url(paste0(urlpath, "20201213_human_KEGG.rda?raw=TRUE")))       # KEGG

The reformatted knowledge-based data were available from the following repositories:

Create a custom-built cell type-related databases by combining different databases for analyzing human single-cell transcriptome data.

d <- list(human_DO[["disease"]], human_CO[["cell"]], human_MSigDB[["cell"]],
          human_CellMarker[["cell"]])
human_CB <- list(cell = do.call("rbind", d))

Add formatted databases to metadata(sce)$sign.

pdacs <- list(CB = pdac, GO = pdac, KG = pdac)
metadata(pdacs$CB) <- list(sign = human_CB[["cell"]])
metadata(pdacs$GO) <- list(sign = human_GO[["BP"]])
metadata(pdacs$KG) <- list(sign = human_KEGG[["pathway"]])

6.3 Create signs

ASURAT function remove_signs() redefines functional gene sets for the input database by removing genes, which are not included in rownames(sce), and further removes biological terms including too few or too many genes.

pdacs$CB <- remove_signs(sce = pdacs$CB, min_ngenes = 2, max_ngenes = 1000)
pdacs$GO <- remove_signs(sce = pdacs$GO, min_ngenes = 2, max_ngenes = 1000)
pdacs$KG <- remove_signs(sce = pdacs$KG, min_ngenes = 2, max_ngenes = 1000)

ASURAT function cluster_genes() clusters functional gene sets using a correlation graph-based decomposition method, which produces strongly, variably, and weakly correlated gene sets (SCG, VCG, and WCG, respectively).

set.seed(1)
pdacs$CB <- cluster_genesets(sce = pdacs$CB, cormat = cormat,
                             th_posi = 0.22, th_nega = -0.32)
set.seed(1)
pdacs$GO <- cluster_genesets(sce = pdacs$GO, cormat = cormat,
                             th_posi = 0.22, th_nega = -0.22)
set.seed(1)
pdacs$KG <- cluster_genesets(sce = pdacs$KG, cormat = cormat,
                             th_posi = 0.18, th_nega = -0.23)

ASURAT function create_signs() creates signs by the following criteria:

the number of genes in SCG>= min_cnt_strg (the default value is 2) and
the number of genes in VCG>= min_cnt_vari (the default value is 2),

which are independently applied to SCGs and VCGs, respectively.

pdacs$CB <- create_signs(sce = pdacs$CB, min_cnt_strg = 2, min_cnt_vari = 2)
pdacs$GO <- create_signs(sce = pdacs$GO, min_cnt_strg = 3, min_cnt_vari = 3)
pdacs$KG <- create_signs(sce = pdacs$KG, min_cnt_strg = 3, min_cnt_vari = 3)

6.4 Select signs

If signs have semantic similarity information, one can use ASURAT function remove_signs_redundant() for removing redundant sings using the semantic similarity matrices.

simmat <- human_GO$similarity_matrix$BP
pdacs$GO <- remove_signs_redundant(sce = pdacs$GO, similarity_matrix = simmat,
                                   threshold = 0.85, keep_rareID = TRUE)

ASURAT function remove_signs_manually() removes signs by specifying IDs (e.g., GOID:XXX) or descriptions (e.g., metabolic) using grepl().

keywords <- "Covid|COVID"
pdacs$KG <- remove_signs_manually(sce = pdacs$KG, keywords = keywords)

6.5 Create sign-by-sample matrices

ASURAT function create_sce_signmatrix() creates a new SingleCellExperiment object new_sce, consisting of the following information:

assayNames(new_sce): counts (SSM whose entries are termed sign scores),
names(colData(new_sce)): nReads, nGenes, percMT,
names(rowData(new_sce)): ParentSignID, Description, CorrGene, etc.,
names(metadata(new_sce)): sign_SCG, sign_VCG, etc.,
altExpNames(new_sce): something if there is data in altExp(sce).

pdacs$CB <- makeSignMatrix(sce = pdacs$CB, weight_strg = 0.5, weight_vari = 0.5)
pdacs$GO <- makeSignMatrix(sce = pdacs$GO, weight_strg = 0.5, weight_vari = 0.5)
pdacs$KG <- makeSignMatrix(sce = pdacs$KG, weight_strg = 0.5, weight_vari = 0.5)

6.6 Reduce dimensions of sign-by-sample matrices

Perform t-distributed stochastic neighbor embedding.

for(i in seq_along(pdacs)){
  set.seed(1)
  mat <- t(as.matrix(assay(pdacs[[i]], "counts")))
  res <- Rtsne::Rtsne(mat, dim = 2, pca = TRUE, initial_dims = 100)
  reducedDim(pdacs[[i]], "TSNE") <- res[["Y"]]
}

Show the results of dimensional reduction in t-SNE spaces.

titles <- c("PDAC-A (cell type & disease)", "PDAC-A (function)",
            "PDAC-A (pathway)")
for(i in seq_along(titles)){
  df <- as.data.frame(reducedDim(pdacs[[i]], "TSNE"))
  p <- ggplot2::ggplot() +
    ggplot2::geom_point(ggplot2::aes(x = df[, 1], y = df[, 2]),
                        color = "black", size = 1, alpha = 1) +
    ggplot2::labs(title = titles[i], x = "tSNE_1", y = "tSNE_2") +
    ggplot2::theme_classic(base_size = 20) +
    ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 18))
  filename <- sprintf("figures/figure_10_%04d.png", 19 + i)
  ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 4.1, height = 4.3)
}

6.7 Cluster cells

6.7.1 Use Seurat functions

To date (December, 2021), one of the most useful clustering methods in scRNA-seq data analysis is a combination of a community detection algorithm and graph-based unsupervised clustering, developed in Seurat package.

Here, our strategy is as follows:

convert SingleCellExperiment objects into Seurat objects (note that rowData() and colData() must have data),
perform ScaleData(), RunPCA(), FindNeighbors(), and FindClusters(),
convert Seurat objects into temporal SingleCellExperiment objects temp,
add colData(temp)$seurat_clusters into colData(sce)$seurat_clusters.

resolutions <- c(0.20, 0.18, 0.15)
dims <- list(seq_len(10), seq_len(30), seq_len(20))
for(i in seq_along(pdacs)){
  surt <- Seurat::as.Seurat(pdacs[[i]], counts = "counts", data = "counts")
  mat <- as.matrix(assay(pdacs[[i]], "counts"))
  surt[["SSM"]] <- Seurat::CreateAssayObject(counts = mat)
  Seurat::DefaultAssay(surt) <- "SSM"
  surt <- Seurat::ScaleData(surt, features = rownames(surt))
  surt <- Seurat::RunPCA(surt, features = rownames(surt))
  surt <- Seurat::FindNeighbors(surt, reduction = "pca", dims = dims[[i]])
  surt <- Seurat::FindClusters(surt, resolution = resolutions[i])
  temp <- Seurat::as.SingleCellExperiment(surt)
  colData(pdacs[[i]])$seurat_clusters <- colData(temp)$seurat_clusters
}

Show the clustering results in t-SNE spaces.

titles <- c("PDAC-A (cell type & disease)", "PDAC-A (function)",
            "PDAC-A (pathway)")
for(i in seq_along(titles)){
  labels <- colData(pdacs[[i]])$seurat_clusters
  df <- as.data.frame(reducedDim(pdacs[[i]], "TSNE"))
  p <- ggplot2::ggplot() +
    ggplot2::geom_point(ggplot2::aes(x = df[, 1], y = df[, 2], color = labels),
                        size = 1, alpha = 1) +
    ggplot2::labs(title = titles[i], x = "tSNE_1", y = "tSNE_2", color = "") +
    ggplot2::theme_classic(base_size = 20) +
    ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 18)) +
    ggplot2::guides(colour = ggplot2::guide_legend(override.aes = list(size = 4)))
  if(i == 1){
    mycolor <- c("0" = "limegreen", "1" = "deepskyblue1", "2" = "grey20",
                 "3" = "red", "4" = "orange")
    p <- p + ggplot2::scale_color_manual(values = mycolor)
  }else if(i == 2){
    mycolor <- c("0" = "limegreen", "1" = "grey20", "2" = "red",
                 "3" = "orange", "4" = "magenta")
    p <- p + ggplot2::scale_color_manual(values = mycolor)
  }else if(i == 3){
    p <- p + ggplot2::scale_color_brewer(palette = "Set2")
  }
  filename <- sprintf("figures/figure_10_%04d.png", 29 + i)
  ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5.1, height = 4.3)
}

Show labels of experimental batches in the reduced sign space.

titles <- c("PDAC-A (cell type & disease)", "PDAC-A (function)",
            "PDAC-A (pathway)")
for(i in seq_along(titles)){
  batch <- as.data.frame(colData(pdacs[[i]])) ; batch$batch <- NA
  batch[grepl("ST", batch$orig.ident), ]$batch <- "ST"
  batch[!grepl("ST", batch$orig.ident), ]$batch <- "inDrop"
  labels <- batch$batch
  df <- as.data.frame(reducedDim(pdacs[[i]], "TSNE"))
  p <- ggplot2::ggplot() +
    ggplot2::geom_point(ggplot2::aes(x = df[, 1], y = df[, 2], color = labels),
                        size = 1, alpha = 1) +
    ggplot2::labs(title = titles[i], x = "tSNE_1", y = "tSNE_2", color = "") +
    ggplot2::theme_classic(base_size = 20) +
    ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 18)) +
    ggplot2::scale_color_hue() +
    ggplot2::guides(colour = ggplot2::guide_legend(override.aes = list(size = 4)))
  filename <- sprintf("figures/figure_10_%04d.png", 34 + i)
  ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5.6, height = 4.3)
}

Show labels of experimental batches of inDrop and ST datasets in the reduced sign space.

# Cell type and disease
batch <- as.character(colData(pdacs$CB)$orig.ident)
labels <- factor(batch, levels = unique(batch))
df <- as.data.frame(reducedDim(pdacs$CB, "TSNE"))
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df[, 1], y = df[, 2], color = labels),
                      size = 1, alpha = 1) +
  ggplot2::labs(title = "PDAC-A (cell type & disease)",
                x = "tSNE_1", y = "tSNE_2", color = "") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 18)) +
  ggplot2::scale_color_hue() +
  ggplot2::guides(colour = ggplot2::guide_legend(override.aes = list(size = 4)))
filename <- "figures/figure_10_0038_a.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 7.1, height = 4.3)

# Biological process
batch <- as.character(colData(pdacs$GO)$orig.ident)
labels <- factor(batch, levels = unique(batch))
df <- as.data.frame(reducedDim(pdacs$GO, "TSNE"))
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df[, 1], y = df[, 2], color = labels),
                      size = 1, alpha = 1) +
  ggplot2::labs(title = "PDAC-A (function)",
                x = "tSNE_1", y = "tSNE_2", color = "") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 18)) +
  ggplot2::scale_color_hue() +
  ggplot2::guides(colour = ggplot2::guide_legend(override.aes = list(size = 4)))
filename <- "figures/figure_10_0038_b.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 7.1, height = 4.3)

Show the number of reads in the reduced sign space.

# Cell type and disease
nreads <- log(as.numeric(colData(pdacs$CB)$nReads) + 1)
df <- as.data.frame(reducedDim(pdacs$CB, "TSNE"))
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df[, 1], y = df[, 2], color = nreads),
                      size = 1, alpha = 1) +
  ggplot2::labs(title = "PDAC-A (cell type & disease)",
                x = "tSNE_1", y = "tSNE_2", color = "log(nReads + 1)") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 18)) +
  ggplot2::scale_color_continuous()
filename <- "figures/figure_10_0039_a.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 6.6, height = 4.3)

# Biological process
nreads <- log(as.numeric(colData(pdacs$GO)$nReads) + 1)
df <- as.data.frame(reducedDim(pdacs$GO, "TSNE"))
p <- ggplot2::ggplot() +
  ggplot2::geom_point(ggplot2::aes(x = df[, 1], y = df[, 2], color = nreads),
                      size = 1, alpha = 1) +
  ggplot2::labs(title = "PDAC-A (function)",
                x = "tSNE_1", y = "tSNE_2", color = "log(nReads + 1)") +
  ggplot2::theme_classic(base_size = 20) +
  ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 18)) +
  ggplot2::scale_color_continuous()
filename <- "figures/figure_10_0039_b.png"
ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 6.6, height = 4.3)

6.8 Investigate significant signs

Significant signs are analogous to differentially expressed genes but bear biological meanings. Note that naïve usages of statistical tests should be avoided because the row vectors of SSMs are centered.

Instead, ASURAT function compute_sepI_all() computes separation indices for each cluster against the others. Briefly, a separation index “sepI”, ranging from -1 to 1, is a nonparametric measure of significance of a given sign score for a given subpopulation. The larger (resp. smaller) sepI is, the more reliable the sign is as a positive (resp. negative) marker for the cluster.

for(i in seq_along(pdacs)){
  set.seed(1)
  labels <- colData(pdacs[[i]])$seurat_clusters
  pdacs[[i]] <- compute_sepI_all(sce = pdacs[[i]], labels = labels,
                                 nrand_samples = NULL)
}

Perform compute_sepI_all() for investigating significant signs for the clustering results of biological process.

labels <- colData(pdacs$GO)$seurat_clusters

pdacs_LabelGO_SignCB <- pdacs$CB
metadata(pdacs_LabelGO_SignCB)$marker_signs <- NULL
set.seed(1)
pdacs_LabelGO_SignCB <- compute_sepI_all(sce = pdacs_LabelGO_SignCB,
                                         labels = labels, nrand_samples = NULL)

pdacs_LabelGO_SignKG <- pdacs$KG
metadata(pdacs_LabelGO_SignKG)$marker_signs <- NULL
set.seed(1)
pdacs_LabelGO_SignKG <- compute_sepI_all(sce = pdacs_LabelGO_SignKG,
                                         labels = labels, nrand_samples = NULL)

6.9 Investigate significant genes

6.9.1 Use Seurat function

To date (December, 2021), one of the most useful methods of multiple statistical tests in scRNA-seq data analysis is to use a Seurat function FindAllMarkers().

If there is gene expression data in altExp(sce), one can investigate differentially expressed genes by using Seurat functions in the similar manner as described before.

set.seed(1)
surt <- Seurat::as.Seurat(pdacs$CB, counts = "counts", data = "counts")
mat <- as.matrix(assay(altExp(pdacs$CB), "counts"))
surt[["GEM"]] <- Seurat::CreateAssayObject(counts = mat)
Seurat::DefaultAssay(surt) <- "GEM"
surt <- Seurat::SetIdent(surt, value = "seurat_clusters")
res <- Seurat::FindAllMarkers(surt, only.pos = TRUE,
                              min.pct = 0.25, logfc.threshold = 0.25)
metadata(pdacs$CB)$marker_genes$all <- res

6.10 Multifaceted analysis

Simultaneously analyze multiple sign-by-sample matrices, which helps us characterize individual samples (cells) from multiple biological aspects.

ASURAT function plot_multiheatmaps() shows heatmaps (ComplexHeatmap object) of sign scores and gene expression levels (if there are), where rows and columns stand for sign (or gene) and sample (cell), respectively.

First, remove unrelated signs by setting keywords, followed by selecting top significant signs and genes for the clustering results with respect to separation index and p-value, respectively.

# Significant signs
marker_signs <- list()
keys <- "foofoo|hogehoge"
for(i in seq_along(pdacs)){
  marker_signs[[i]] <- metadata(pdacs[[i]])$marker_signs$all
  marker_signs[[i]] <- marker_signs[[i]][!grepl(keys, marker_signs[[i]]$Description), ]
  marker_signs[[i]] <- dplyr::group_by(marker_signs[[i]], Ident_1)
  marker_signs[[i]] <- dplyr::slice_max(marker_signs[[i]], sepI, n = 2)
  marker_signs[[i]] <- dplyr::slice_min(marker_signs[[i]], Rank, n = 1)
}
# Significant genes
marker_genes_CB <- metadata(pdacs$CB)$marker_genes$all
marker_genes_CB <- dplyr::group_by(marker_genes_CB, cluster)
marker_genes_CB <- dplyr::slice_min(marker_genes_CB, p_val_adj, n = 2)
marker_genes_CB <- dplyr::slice_max(marker_genes_CB, avg_log2FC, n = 2)

Then, prepare arguments.

# ssm_list
sces_sub <- list() ; ssm_list <- list()
for(i in seq_along(pdacs)){
  sces_sub[[i]] <- pdacs[[i]][rownames(pdacs[[i]]) %in% marker_signs[[i]]$SignID, ]
  ssm_list[[i]] <- assay(sces_sub[[i]], "counts")
}
names(ssm_list) <- c("SSM_cell_disease", "SSM_function", "SSM_pathway")
# gem_list
expr_sub <- altExp(pdacs$CB, "logcounts")
expr_sub <- expr_sub[rownames(expr_sub) %in% marker_genes_CB$gene]
gem_list <- list(x = t(scale(t(as.matrix(assay(expr_sub, "counts"))))))
names(gem_list) <- "Scaled\nLogExpr"
# ssmlabel_list
labels <- list() ; ssmlabel_list <- list()
for(i in seq_along(pdacs)){
  tmp <- colData(sces_sub[[i]])$seurat_clusters
  labels[[i]] <- data.frame(label = tmp)
  n_groups <- length(unique(tmp))
  if(i == 1){
    mycolor <- c("0" = "limegreen", "1" = "deepskyblue1", "2" = "grey20",
                 "3" = "red", "4" = "orange")
    labels[[i]]$color <- mycolor[tmp]
  }else if(i == 2){
    mycolor <- c("0" = "limegreen", "1" = "grey20", "2" = "red",
                 "3" = "orange", "4" = "magenta")
    labels[[i]]$color <- mycolor[tmp]
  }else if(i == 3){
    labels[[i]]$color <- scales::brewer_pal(palette = "Set2")(n_groups)[tmp]
  }
  ssmlabel_list[[i]] <- labels[[i]]
}
names(ssmlabel_list) <- c("Label_cell_disease", "Label_function", "Label_pathway")

Finally, plot heatmaps for the selected signs and genes.

filename <- "figures/figure_10_0040.png"
# png(file = filename, height = 1200, width = 1300, res = 200)
png(file = filename, height = 350, width = 370, res = 60)
set.seed(1)
title <- "PDAC-A"
plot_multiheatmaps(ssm_list = ssm_list, gem_list = gem_list,
                   ssmlabel_list = ssmlabel_list, gemlabel_list = NULL,
                   nrand_samples = 500, show_row_names = TRUE, title = title)
dev.off()

Show violin plots for sign score and gene expression distributions across cell type-related clusters.

labels <- colData(pdacs$CB)$seurat_clusters
vlist <- list(c("GE", "REG1A", "(Pancreatic cell)\n"),
              c("GE", "CPA5", "(Pancreatic cell)\n"),
              c("CB", "MSigDBID:252-V",
                "...PANCREAS_DUCTAL_CELL\n(SOX9, APOL1, ...)"),
              c("CB", "DOID:3498-S",
                "pancreatic ductal adenocarcinoma\n(S100P, MMP9, ...)"),
              c("CB", "MSigDBID:69-S", "...NK_CELLS\n(GZMB, HCST, ...)"),
              c("CB", "MSigDBID:41-S", "...MACROPHAGE\n(TYROBP, MS4A7, ...)"))
ylabels <- list(GE = "Expression level", CB = "Sign score")
mycolor <- c("0" = "limegreen", "1" = "deepskyblue1", "2" = "grey20", "3" = "red",
             "4" = "orange")
for(i in seq_along(vlist)){
  if(vlist[[i]][1] == "GE"){
    ind <- which(rownames(altExp(pdacs$CB, "logcounts")) == vlist[[i]][2])
    subsce <- altExp(pdacs$CB, "logcounts")[ind, ]
    df <- as.data.frame(t(as.matrix(assay(subsce, "counts"))))
  }else{
    ind <- which(rownames(pdacs[[vlist[[i]][1]]]) == vlist[[i]][2])
    subsce <- pdacs[[vlist[[i]][1]]][ind, ]
    df <- as.data.frame(t(as.matrix(assay(subsce, "counts"))))
  }
  p <- ggplot2::ggplot() +
    ggplot2::geom_violin(ggplot2::aes(x = as.factor(labels), y = df[, 1],
                                      fill = labels), trim = FALSE, size = 0.5) +
    ggplot2::geom_boxplot(ggplot2::aes(x = as.factor(labels), y = df[, 1]),
                          width = 0.15, alpha = 0.6) +
    ggplot2::labs(title = paste0(vlist[[i]][2], "\n", vlist[[i]][3]),
                  x = "Cluster (cell type & disease)",
                  y = ylabels[[vlist[[i]][1]]]) +
    ggplot2::theme_classic(base_size = 25) +
    ggplot2::theme(plot.title = ggplot2::element_text(hjust = 0.5, size = 20),
                   legend.position = "none") +
    ggplot2::scale_fill_manual(values = mycolor)
  filename <- sprintf("figures/figure_10_%04d.png", 49 + i)
  ggplot2::ggsave(file = filename, plot = p, dpi = 50, width = 5, height = 4)
}

Analysis of PDAC datasets (Moncada et al., 2020)

Keita Iida

2022-09-07

1 Computational environment

2 Install libraries

3 Introduction