1. singleCellTK
  2. Commit 2d8ecbf8db4170ebc9f2544e062a816713ddeb02
Browse code

New function: diffAbundanceFET, plotClusterAbundance

Yichen Wang authored on 13/05/2021 21:09:54
Showing 5 changed files
NAMESPACE
History View file @ 2d8ecbf8
... ... 
@@ -10,6 +10,7 @@ export(convertGeneIDs)
10 10 
 export(convertSCEToSeurat)
11 11 
 export(convertSeuratToSCE)
12 12 
 export(dedupRowNames)
13 
+export(diffAbundanceFET)
13 14 
 export(discreteColorPalette)
14 15 
 export(distinctColors)
15 16 
 export(downSampleCells)
... ... 
@@ -63,6 +64,7 @@ export(plotBarcodeRankScatter)
63 64 
 export(plotBatchVariance)
64 65 
 export(plotBcdsResults)
65 66 
 export(plotBiomarker)
67 
+export(plotClusterAbundance)
66 68 
 export(plotCxdsResults)
67 69 
 export(plotDEGHeatmap)
68 70 
 export(plotDEGRegression)
R/abundance.R
History View file @ 2d8ecbf8
69 71 
new file mode 100644
... ... 
@@ -0,0 +1,173 @@
1 
+#' Calculate Differential Abundance with FET
2 
+#' @details This function will calculate the cell counting and fraction by
3 
+#' dividing all cells to groups specified by the arguments, together with
4 
+#' statistical summary by performing Fisher Exact Tests (FET).
5 
+#' @param inSCE A \code{\link[SingleCellExperiment]{SingleCellExperiment}}
6 
+#' object.
7 
+#' @param cluster A single \code{character}, specifying the name to store the
8 
+#' cluster label in \code{\link{colData}}.
9 
+#' @param variable A single \code{character}, specifying the name to store the
10 
+#' phenotype labels in \code{\link{colData}}.
11 
+#' @param control \code{character}. Specifying one or more categories that can
12 
+#' be found in the vector specified by \code{variable}.
13 
+#' @param case \code{character}. Specifying one or more categories that can
14 
+#' be found in the vector specified by \code{variable}.
15 
+#' @param analysisName A single \code{character}. Will be used for naming the
16 
+#' result table, which will be saved in metadata slot.
17 
+#' @return The original \code{\link[SingleCellExperiment]{SingleCellExperiment}}
18 
+#' object with \code{metadata(inSCE)} updated with a list
19 
+#' \code{diffAbundanceFET}, containing a new \code{data.frame} for the analysis
20 
+#' result, named by \code{analysisName}. The \code{data.frame} contains columns
21 
+#' for number and fraction of cells that belong to different cases, as well as
22 
+#' "Odds_Ratio", "PValue" and "FDR".
23 
+#' @export
24 
+#' @examples
25 
+#' data("mouseBrainSubsetSCE", package = "singleCellTK")
26 
+#' mouseBrainSubsetSCE <- diffAbundanceFET(inSCE = mouseBrainSubsetSCE,
27 
+#'                                                 cluster = "tissue",
28 
+#'                                                 variable = "level1class",
29 
+#'                                                 case = "oligodendrocytes",
30 
+#'                                                 control = "microglia",
31 
+#'                                                 analysisName = "diffAbundFET")
32 
+diffAbundanceFET <- function(inSCE, cluster, variable, control, case,
33 
+                             analysisName) {
34 
+  if (!inherits(inSCE, "SingleCellExperiment")) {
35 
+    stop("'inSCE' should be a SingleCellExperiment object.")
36 
+  }
37 
+  if (is.null(cluster) ||
38 
+      !cluster %in% names(SummarizedExperiment::colData(inSCE))) {
39 
+    stop("Argument 'cluster' should be a column name in colData(inSCE).")
40 
+  }
41 
+  if (is.null(variable) ||
42 
+      !variable %in% names(SummarizedExperiment::colData(inSCE))) {
43 
+    stop("Argument 'variable' should be a column name in colData(inSCE).")
44 
+  }
45 
+
46 
+  cluster <- SummarizedExperiment::colData(inSCE)[, cluster]
47 
+  variable <- SummarizedExperiment::colData(inSCE)[, variable]
48 
+  if (!all(control %in% variable)) {
49 
+    nf1 <- control[which(!control %in% variable)]
50 
+    stop("Given 'control' not all found in 'variable': ",
51 
+         paste(nf1, collapse = ", "))
52 
+  }
53 
+  if (!all(case %in% variable)) {
54 
+    nf2 <- case[which(!case %in% variable)]
55 
+    stop("Given 'case' not all found in 'variable': ",
56 
+         paste(nf2, collapse = ", "))
57 
+  }
58 
+  if (any(control %in% case)) {
59 
+    warning("Overlapping variables found in 'control' and 'case'")
60 
+  }
61 
+
62 
+  control.lab <- paste(control, collapse=",")
63 
+  case.lab <- paste(case, collapse=",")
64 
+  label <- rep(NA, length(variable))
65 
+  label[variable %in% case] <- case.lab
66 
+  label[variable %in% control] <- control.lab
67 
+  label <- factor(label, levels=c(control.lab, case.lab))
68 
+
69 
+  res <- matrix(NA, nrow=length(unique(cluster)), ncol=10)
70 
+  cluster.label <- sort(unique(cluster))
71 
+  for(i in seq_along(cluster.label)) {
72 
+    cluster.factor <- factor(ifelse(cluster %in% cluster.label[i],
73 
+                                    cluster.label[i], "Other"),
74 
+                             levels=c(i, "Other"))
75 
+
76 
+    ta <- table(cluster.factor, label)
77 
+    ta.p <- prop.table(ta, 2)
78 
+    fet <- stats::fisher.test(ta)
79 
+    res[i,] <- c(ta, ta.p, fet$estimate, fet$p.value)
80 
+  }
81 
+  colnames(res) <- c(paste0("Number of cells in cluster and in ",
82 
+                            control.lab),
83 
+                     paste0("Number of cells NOT in cluster and in ",
84 
+                            control.lab),
85 
+                     paste0("Number of cells in cluster and in ",
86 
+                            case.lab),
87 
+                     paste0("Number of cells NOT in cluster and in ",
88 
+                            case.lab),
89 
+                     paste0("Fraction of cells in cluster and in ",
90 
+                            control.lab),
91 
+                     paste0("Fraction of cells NOT in cluster and in ",
92 
+                            control.lab),
93 
+                     paste0("Fraction of cells in cluster and in ",
94 
+                            case.lab),
95 
+                     paste0("Fraction of cells NOT in cluster and in ",
96 
+                            case.lab),
97 
+                     "Odds_Ratio", "Pvalue")
98 
+  res <- data.frame(Cluster=cluster.label, res,
99 
+                    FDR=stats::p.adjust(res[,"Pvalue"], 'fdr'),
100 
+                    check.names=FALSE)
101 
+  if (!"diffAbundanceFET" %in% names(S4Vectors::metadata(inSCE))) {
102 
+    S4Vectors::metadata(inSCE)$diffAbundanceFET <- list()
103 
+  }
104 
+  S4Vectors::metadata(inSCE)$diffAbundanceFET[[analysisName]] <- res
105 
+
106 
+  return(inSCE)
107 
+}
108 
+
109 
+#' Plot the differential Abundance
110 
+#' @details This function will visualize the differential abundance in two given
111 
+#' variables, by making bar plots that presents the cell counting and fraction
112 
+#' in different cases.
113 
+#' @param inSCE A \code{\link[SingleCellExperiment]{SingleCellExperiment}}
114 
+#' object.
115 
+#' @param cluster A single \code{character}, specifying the name to store the
116 
+#' cluster label in \code{\link{colData}}.
117 
+#' @param variable A single \code{character}, specifying the name to store the
118 
+#' phenotype labels in \code{\link{colData}}.
119 
+#' @return A \code{list} with 4 \code{ggplot} objects.
120 
+#' @export
121 
+#' @examples
122 
+#' data("mouseBrainSubsetSCE", package = "singleCellTK")
123 
+#' plotClusterAbundance(inSCE = mouseBrainSubsetSCE,
124 
+#'                      cluster = "tissue",
125 
+#'                      variable = "level1class")
126 
+plotClusterAbundance <- function(inSCE, cluster, variable) {
127 
+  if (!inherits(inSCE, "SingleCellExperiment")) {
128 
+    stop("'inSCE' should be a SingleCellExperiment object.")
129 
+  }
130 
+  if (is.null(cluster) ||
131 
+      !cluster %in% names(SummarizedExperiment::colData(inSCE))) {
132 
+    stop("Argument 'cluster' should be a column name in colData(inSCE).")
133 
+  }
134 
+  if (is.null(variable) ||
135 
+      !variable %in% names(SummarizedExperiment::colData(inSCE))) {
136 
+    stop("Argument 'variable' should be a column name in colData(inSCE).")
137 
+  }
138 
+  cluster <- SummarizedExperiment::colData(inSCE)[, cluster]
139 
+  color_palette <- distinctColors(length(unique(cluster)))
140 
+
141 
+  label <- SummarizedExperiment::colData(inSCE)[, variable]
142 
+
143 
+  cluster.color <- color_palette
144 
+  df <- data.frame(Cluster=as.factor(cluster), Sample=as.factor(label))
145 
+
146 
+  g1 <- ggplot2::ggplot(df, ggplot2::aes_string("Cluster")) +
147 
+    ggplot2::geom_bar(ggplot2::aes_string(fill="Sample")) +
148 
+    ggplot2::theme_classic() +
149 
+    ggplot2::scale_y_continuous(expand = c(0, 0))
150 
+
151 
+  g2 <- ggplot2::ggplot(df, ggplot2::aes_string("Cluster")) +
152 
+    ggplot2::geom_bar(ggplot2::aes_string(fill="Sample"), position="fill") +
153 
+    ggplot2::theme_classic() +
154 
+    ggplot2::scale_y_continuous(expand = c(0, 0)) +
155 
+    ggplot2:: ylab("Fraction")
156 
+
157 
+  g3 <- ggplot2::ggplot(df, ggplot2::aes_string("Sample")) +
158 
+    ggplot2::geom_bar(ggplot2::aes_string(fill="Cluster")) +
159 
+    ggplot2::theme_classic() +
160 
+    ggplot2::scale_y_continuous(expand = c(0, 0)) +
161 
+    ggplot2::theme(axis.text.x = ggplot2::element_text(angle = 45, hjust = 1)) +
162 
+    ggplot2::scale_fill_manual(values=cluster.color)
163 
+
164 
+  g4 <- ggplot2::ggplot(df, ggplot2::aes_string("Sample")) +
165 
+    ggplot2::geom_bar(ggplot2::aes_string(fill="Cluster"), position="fill") +
166 
+    ggplot2::theme_classic() +
167 
+    ggplot2::scale_y_continuous(expand = c(0, 0)) +
168 
+    ggplot2::theme(axis.text.x = ggplot2::element_text(angle = 45, hjust = 1)) +
169 
+    ggplot2::scale_fill_manual(values=cluster.color) +
170 
+    ggplot2::ylab("Fraction")
171 
+  g <- list(g1, g2, g3, g4)
172 
+  return(g)
173 
+}
0 174 
\ No newline at end of file
R/getCluster.R
History View file @ 2d8ecbf8
... ... 
@@ -58,7 +58,7 @@ runScranSNN <- function(inSCE, useAssay = NULL, useReducedDim = NULL,
58 58 
                                       "leadingEigen")) {
59 59 
   # TODO: parallele parameter
60 60 
   if (!inherits(inSCE, "SingleCellExperiment")) {
61 
-    stop("SCE")
61 
+    stop("'inSCE' should be a SingleCellExperiment object.")
62 62 
   }
63 63 
   if (is.null(useAssay) + is.null(useReducedDim) + is.null(useAltExp) != 2) {
64 64 
     stop("Scran SNN clustering requires one and only one of 'useAssay', ",
... ... 
@@ -148,7 +148,7 @@ runKMeans <- function(inSCE, useReducedDim = "PCA",
148 148 
                       nStart = 1, seed = 12345,
149 149 
                       algorithm = c("Hartigan-Wong", "Lloyd", "MacQueen")){
150 150 
   if (!inherits(inSCE, "SingleCellExperiment")) {
151 
-    stop("SCE")
151 
+    stop("'inSCE' should be a SingleCellExperiment object.")
152 152 
   }
153 153 
   if (is.null(useReducedDim)) {
154 154 
     stop("runKMeans requires 'useReducedDim' input.")
man/diffAbundanceFET.Rd
History View file @ 2d8ecbf8
155 155 
new file mode 100644
... ... 
@@ -0,0 +1,52 @@
1 
+% Generated by roxygen2: do not edit by hand
2 
+% Please edit documentation in R/abundance.R
3 
+\name{diffAbundanceFET}
4 
+\alias{diffAbundanceFET}
5 
+\title{Calculate Differential Abundance with FET}
6 
+\usage{
7 
+diffAbundanceFET(inSCE, cluster, variable, control, case, analysisName)
8 
+}
9 
+\arguments{
10 
+\item{inSCE}{A \code{\link[SingleCellExperiment]{SingleCellExperiment}}
11 
+object.}
12 
+
13 
+\item{cluster}{A single \code{character}, specifying the name to store the
14 
+cluster label in \code{\link{colData}}.}
15 
+
16 
+\item{variable}{A single \code{character}, specifying the name to store the
17 
+phenotype labels in \code{\link{colData}}.}
18 
+
19 
+\item{control}{\code{character}. Specifying one or more categories that can
20 
+be found in the vector specified by \code{variable}.}
21 
+
22 
+\item{case}{\code{character}. Specifying one or more categories that can
23 
+be found in the vector specified by \code{variable}.}
24 
+
25 
+\item{analysisName}{A single \code{character}. Will be used for naming the
26 
+result table, which will be saved in metadata slot.}
27 
+}
28 
+\value{
29 
+The original \code{\link[SingleCellExperiment]{SingleCellExperiment}}
30 
+object with \code{metadata(inSCE)} updated with a list
31 
+\code{diffAbundanceFET}, containing a new \code{data.frame} for the analysis
32 
+result, named by \code{analysisName}. The \code{data.frame} contains columns
33 
+for number and fraction of cells that belong to different cases, as well as
34 
+"Odds_Ratio", "PValue" and "FDR".
35 
+}
36 
+\description{
37 
+Calculate Differential Abundance with FET
38 
+}
39 
+\details{
40 
+This function will calculate the cell counting and fraction by
41 
+dividing all cells to groups specified by the arguments, together with
42 
+statistical summary by performing Fisher Exact Tests (FET).
43 
+}
44 
+\examples{
45 
+data("mouseBrainSubsetSCE", package = "singleCellTK")
46 
+mouseBrainSubsetSCE <- diffAbundanceFET(inSCE = mouseBrainSubsetSCE,
47 
+                                                cluster = "tissue",
48 
+                                                variable = "level1class",
49 
+                                                case = "oligodendrocytes",
50 
+                                                control = "microglia",
51 
+                                                analysisName = "diffAbundFET")
52 
+}
man/plotClusterAbundance.Rd
History View file @ 2d8ecbf8
0 53 
new file mode 100644
... ... 
@@ -0,0 +1,35 @@
1 
+% Generated by roxygen2: do not edit by hand
2 
+% Please edit documentation in R/abundance.R
3 
+\name{plotClusterAbundance}
4 
+\alias{plotClusterAbundance}
5 
+\title{Plot the differential Abundance}
6 
+\usage{
7 
+plotClusterAbundance(inSCE, cluster, variable)
8 
+}
9 
+\arguments{
10 
+\item{inSCE}{A \code{\link[SingleCellExperiment]{SingleCellExperiment}}
11 
+object.}
12 
+
13 
+\item{cluster}{A single \code{character}, specifying the name to store the
14 
+cluster label in \code{\link{colData}}.}
15 
+
16 
+\item{variable}{A single \code{character}, specifying the name to store the
17 
+phenotype labels in \code{\link{colData}}.}
18 
+}
19 
+\value{
20 
+A \code{list} with 4 \code{ggplot} objects.
21 
+}
22 
+\description{
23 
+Plot the differential Abundance
24 
+}
25 
+\details{
26 
+This function will visualize the differential abundance in two given
27 
+variables, by making bar plots that presents the cell counting and fraction
28 
+in different cases.
29 
+}
30 
+\examples{
31 
+data("mouseBrainSubsetSCE", package = "singleCellTK")
32 
+plotClusterAbundance(inSCE = mouseBrainSubsetSCE,
33 
+                     cluster = "tissue",
34 
+                     variable = "level1class")
35 
+}