replace avg_logFC to avg_log2FC in multiple files

R/cellbrowser_prep.R

@@ -102,14 +102,14 @@ print(file.path(opt$seurat_path, opt$runspecs,
                 "cluster.markers.dir","markers.summary.table.xlsx"))
 data_selected = openxlsx::read.xlsx(xlsxFile = file.path(opt$seurat_path, opt$runspecs,
                                               "cluster.markers.dir","markers.summary.table.xlsx"))
-output = data_selected[,c("gene","gene_id", "cluster","avg_logFC","p.adj")]
-output = output %>% group_by(cluster) %>% dplyr::arrange(desc(avg_logFC)) %>% do(head(.,n=20)) %>% ungroup()
+output = data_selected[,c("gene","gene_id", "cluster","avg_log2FC","p.adj")]
+output = output %>% group_by(cluster) %>% dplyr::arrange(desc(avg_log2FC)) %>% do(head(.,n=20)) %>% ungroup()
 output$celltype = "top20_marker_Seurat_cluster"
 output$cluster_marker = paste(output$celltype, output$cluster, sep="_")
 output = output %>% dplyr::select(-celltype)
-colnames(output) = c("gene","gene_id","cluster","avg_logFC","p_adjusted","celltype_marker")
-output = output[,c("cluster","gene","p_adjusted","avg_logFC","celltype_marker")]
+colnames(output) = c("gene","gene_id","cluster","avg_log2FC","p_adjusted","celltype_marker")
+output = output[,c("cluster","gene","p_adjusted","avg_log2FC","celltype_marker")]
 write.table(output, file.path(opt$outdir, "markers.tsv"), sep = "\t",
           quote = FALSE, row.names = FALSE)
 
-cat("Completed")
+cat("Completed")

R/genesetAnalysis.R

@@ -64,10 +64,10 @@ cluster_de <- de[de$cluster==opt$cluster & de$p.adj<=opt$adjpthreshold,]
 ## only positive markers considered
 if(opt$direction=="positive")
 {
-    foreground <- unique(cluster_de$gene_id[cluster_de$avg_logFC>0])
+    foreground <- unique(cluster_de$gene_id[cluster_de$avg_log2FC>0])
 } else if (opt$direction=="negative")
 {
-    foreground <- unique(cluster_de$gene_id[cluster_de$avg_logFC<0])
+    foreground <- unique(cluster_de$gene_id[cluster_de$avg_log2FC<0])
 } else if (opt$direction=="both")
 {
     foreground <- unique(cluster_de$gene_id)

R/seurat_FindMarkers.R

@@ -297,7 +297,7 @@ for (conserved.level in levels(ident.conserved)){
         return.thresh = 1
 
         if (nrow(markers) > 0) {
-            markers = markers[order(markers$p_val, -markers$avg_logFC), ]
+            markers = markers[order(markers$p_val, -markers$avg_log2FC), ]
             markers = subset(markers, p_val < return.thresh)
 
             if (nrow(markers) > 0){
@@ -314,7 +314,7 @@ for (conserved.level in levels(ident.conserved)){
         markers$p.adj <- p.adjust(markers$p_val, method="BH")
 
         message("selecting columns of interest")
-        markers <- markers[,c("cluster","gene","p.adj","p_val","avg_logFC","pct.1","pct.2")]
+        markers <- markers[,c("cluster","gene","p.adj","p_val","avg_log2FC","pct.1","pct.2")]
         markers <- markers[order(markers$cluster, markers$p_val),]
 
         print(head(markers))
@@ -425,7 +425,7 @@ if (length(markers.conserved.list) > 1){
     table(markers.sig)
     markers.sig <- names(markers.sig)[markers.sig]
 
-    markers.fc <- do.call("cbind", lapply(markers.conserved.list, function(x){x[markers.sig, "avg_logFC"]}))
+    markers.fc <- do.call("cbind", lapply(markers.conserved.list, function(x){x[markers.sig, "avg_log2FC"]}))
     rownames(markers.fc) <- markers.sig
     colnames(markers.fc) <- names(markers.conserved.list)
 
@@ -480,14 +480,14 @@ if (length(markers.conserved.list) > 1){
         tmp.table <- do.call(
           "cbind",
           lapply(markers.conserved.list, function(x){
-            x[markers.conserved, "avg_logFC"]
+            x[markers.conserved, "avg_log2FC"]
           })
         )
         rownames(tmp.table) <- markers.conserved
         tmp.table <- exp(tmp.table) - 1
-        conserved.table$avg_logFC <- log(rowMeans(as.matrix(tmp.table[markers.conserved, , drop=FALSE])) + 1)
+        conserved.table$avg_log2FC <- log(rowMeans(as.matrix(tmp.table[markers.conserved, , drop=FALSE])) + 1)
 
-        message("avg_logFC added")
+        message("avg_log2FC added")
 
         # average percentage of cells in group 1 ----
         tmp.table <- do.call(
@@ -547,7 +547,7 @@ if (length(markers.conserved.list) > 1){
             gene_id = character(0),
             p_val = character(0),
             p.adj = character(0),
-            avg_logFC = character(0),
+            avg_log2FC = character(0),
             pct.1 = character(0),
             pct.2 = character(0),
             cluster_mean = character(0),

R/seurat_cluster_marker_plots.R

@@ -112,13 +112,13 @@ x <- x[x$cluster==as.numeric(opt$cluster) & x$p.adj<0.1,]
 
 n_select = 16
 # pull out by average and minimum log fold change
-x %>% top_n(n_select, avg_logFC) -> top_by_avg_logFC
+x %>% top_n(n_select, avg_log2FC) -> top_by_avg_log2FC
 
 n_select = 8
-x[!x$gene %in% top_by_avg_logFC$gene,] %>%
+x[!x$gene %in% top_by_avg_log2FC$gene,] %>%
   top_n(n_select, min_logFC) -> top_by_min_logFC
 
-x <- x[x$gene %in% c(top_by_avg_logFC$gene,
+x <- x[x$gene %in% c(top_by_avg_log2FC$gene,
                      top_by_min_logFC$gene),]
 
 # marker gene heatmap.

R/seurat_summariseMarkerNumbers.R

@@ -81,8 +81,8 @@ message("summarising and melting the data")
 summarised_data <- degenes %>%
     dplyr::group_by(cluster) %>%
     dplyr::summarise(
-      positive=length(which(p.adj < opt$minpadj & avg_logFC >= log(opt$minfc))),
-      negative=length(which(p.adj < opt$minpadj & avg_logFC <= -log(opt$minfc))))
+      positive=length(which(p.adj < opt$minpadj & avg_log2FC >= log(opt$minfc))),
+      negative=length(which(p.adj < opt$minpadj & avg_log2FC <= -log(opt$minfc))))
 
 melted_data <- melt(summarised_data, id=c("cluster"))
 

R/seurat_summariseMarkers.R

@@ -93,7 +93,7 @@ print(dim(markers))
 
 markers <- markers[,c("cluster","gene","gene_id",
                       "p_val","p.adj",
-                      "avg_logFC","pct.1","pct.2",
+                      "avg_log2FC","pct.1","pct.2",
                       "cluster_mean","other_mean")]
 
 markers <- markers[order(markers$cluster, markers$p.adj),]
@@ -235,7 +235,7 @@ saveWorkbook(wb, file=paste(outPrefix,"table","xlsx",
 message("Making a heatmap of the top marker genes from each cluster")
 
 ## make a heatmap of the top DE genes.
-filtered_markers %>% group_by(cluster) %>% top_n(20, avg_logFC) -> top20
+filtered_markers %>% group_by(cluster) %>% top_n(20, avg_log2FC) -> top20
 
 if(!is.null(opt$subgroup))
 {
@@ -287,8 +287,8 @@ summary <- c()
 for(id in idents.all)
 {
     ncells = length(cluster_ids[cluster_ids==id])
-    npos = length(filtered_markers$p.adj[filtered_markers$cluster==id & filtered_markers$avg_logFC > 0] )
-    nneg = length(filtered_markers$p.adj[filtered_markers$cluster==id & filtered_markers$avg_logFC < 0] )
+    npos = length(filtered_markers$p.adj[filtered_markers$cluster==id & filtered_markers$avg_log2FC > 0] )
+    nneg = length(filtered_markers$p.adj[filtered_markers$cluster==id & filtered_markers$avg_log2FC < 0] )
     ntotal = npos + nneg
     summary <- c(summary,c(id, ncells, npos, nneg, ntotal))
 }

R/seurat_summariseMarkersBetween.R

@@ -126,7 +126,7 @@ for(cluster in clusters)
        }
 }
 
-res <- res[,c("cluster","gene","p.adj","p_val","avg_logFC",
+res <- res[,c("cluster","gene","p.adj","p_val","avg_log2FC",
               "pct.1","pct.2",aName,bName,"gene_id")]
 
 out_fn <- file.path(
@@ -222,7 +222,7 @@ save_ggplots(gsub(".tex",".nde",tex_fn),
              height=5)
 
 ## Make a heatmap
-diffMat <- dcast(res, gene~cluster, value.var="avg_logFC")
+diffMat <- dcast(res, gene~cluster, value.var="avg_log2FC")
 diffMat[is.na(diffMat)] <- 0
 
 rownames(diffMat) <- diffMat$gene
@@ -258,7 +258,7 @@ plot_fn <- function()
                   density.info=c("none"),
                   lwid = c(1,5),
                   lhei = c(1,8),
-                  key.xlab = "avg_logFC",
+                  key.xlab = "avg_log2FC",
                   key.ylab = "",
                   xlab="cluster",
                   cexRow = 0.4,

R/seurat_topMarkerHeatmap.R

@@ -59,7 +59,7 @@ markers <- read.table(opt$markers,
 filtered_markers <- data.table(markers[markers$p.adj < 0.1,])
 
 ## make a heatmap of the top DE genes.
-filtered_markers %>% group_by(cluster) %>% top_n(20, avg_logFC) -> top20
+filtered_markers %>% group_by(cluster) %>% top_n(20, avg_log2FC) -> top20
 
 if(!is.null(opt$subgroup))
 {

pipelines/pipeline_scxl.py

@@ -2443,7 +2443,7 @@ def topClusterMarkers(infile, outfile):
     def _filterAndScore(data):
         # filter for strong cluster markers
         data = data[(data["p.adj"] < 0.01) &
-                    (data["avg_logFC"].abs() > np.log(2)) &
+                    (data["avg_log2FC"].abs() > np.log(2)) &
                     (data["cluster_mean"] > 2) &
                     (data["pct.1"] > 0.25)]
 
@@ -2452,7 +2452,7 @@ def _filterAndScore(data):
         # for fold change, expression level and adjusted p-value.
         # the aim is to give "better" markers higher scores.
         pscore = [1 - x for x in data["p.adj"].values]
-        fscore = [np.exp(np.abs(x)) for x in data["avg_logFC"].values]
+        fscore = [np.exp(np.abs(x)) for x in data["avg_log2FC"].values]
         escore = [np.log2(x) for x in data["cluster_mean"].values]
 
         # construct a matrix of the scores and take the geometric mean.
@@ -2535,7 +2535,7 @@ def _write_tables(d, name="none"):
 
     # keep up to n entries per cluster
     # note that groupby preserves the ordering.
-    positive_markers = data[data["avg_logFC"] > 0]
+    positive_markers = data[data["avg_log2FC"] > 0]
     positive_markers = _filterAndScore(positive_markers)
     positive_markers = _skimMarkers(positive_markers,
                                     PARAMS["exprsreport_n_positive"])
@@ -2547,7 +2547,7 @@ def _write_tables(d, name="none"):
     if stat:
         statements.append(stat)
 
-    negative_markers = data[data["avg_logFC"] < 0]
+    negative_markers = data[data["avg_log2FC"] < 0]
     negative_markers = _filterAndScore(negative_markers)
     negative_markers = _skimMarkers(negative_markers,
                                     PARAMS["exprsreport_n_negative"])

python/make_anndata.py

@@ -5,7 +5,7 @@
 import pandas as pd
 import logging
 import sys
-
+import hnswlib
 
 
 # ########################################################################### #

python/run_paga.py

@@ -14,7 +14,7 @@
 from scipy import sparse
 import logging
 import sys
-
+import igraph
 
 
 # ########################################################################### #

tenxutils/R/Helper.R

@@ -52,7 +52,7 @@ sprintfResults <- function(results_table,
 #' @param m_col The column containing the log2 ratio
 #' @param id_col A column containing unique identifiers
 #' @param ngenes The number of genes to demarcate
-topGenes <- function(data, m_col = "avg_logFC",
+topGenes <- function(data, m_col = "avg_log2FC",
                      use_fc = TRUE,
                      id_col="gene", ngenes=7)
 {
@@ -76,7 +76,7 @@ topGenes <- function(data, m_col = "avg_logFC",
 #' @param m_col The column containing the log2 ratio
 #' @param ngenes The number of genes to demarcate
 #' @param id_col A column containing a unique identifier
-categoriseGenes <- function(data,m_col="avg_logFC", use_fc=TRUE,
+categoriseGenes <- function(data,m_col="avg_log2FC", use_fc=TRUE,
                             p_col="p.adj", p_threshold=0.05,
                             ngenes=7,
                             id_col="gene")

tenxutils/R/Plot.R

@@ -134,8 +134,8 @@ plotViolins <- function(data, seurat_object,
                         vncol=4, vnrow=3,
                         use.minfc=FALSE,
                         minfc_col="min_logFC", maxfc_col="max_logFC",
-                        avgfc_col="avg_logFC",
-                        m_col="avg_logFC", p_col="p.adj",
+                        avgfc_col="avg_log2FC",
+                        m_col="avg_log2FC", p_col="p.adj",
                         pt_size=0.1,
                         id_col="gene")
 {
@@ -206,7 +206,7 @@ plotViolins <- function(data, seurat_object,
         tmp <- tmp[!tmp[[id_col]] %in% genes_a,]
 
         ## order by largest fold change
-        ## tmp <- tmp[rev(order(abs(tmp$avg_logFC))),]
+        ## tmp <- tmp[rev(order(abs(tmp$avg_log2FC))),]
 
         ## subset to positive or negative markers
         if(type == "positive")
@@ -545,7 +545,7 @@ plotDownsampling <- function(matrixUMI, metadata, basename) {
 #' A function to draw a heatmap of top cluster marker genes with
 #' subgroup labels. The function uses the "scale.data" slot by default to make the heatmap.
 #' @param seurat_object A seurat objected with scaled data and cluster information
-#' @param marker_table A dataframe containing the marker information. Must contain "cluster", "gene" and "avg_logFC" columns
+#' @param marker_table A dataframe containing the marker information. Must contain "cluster", "gene" and "avg_log2FC" columns
 #' @param n_markers The number of markers to plot
 #' @param cells_use The names of the cells to use. If NULL all cells will be used
 #' @param row_names_gp The font size for the gene names
@@ -558,7 +558,7 @@ markerComplexHeatmap <- function(seurat_object,
                                  n_markers=20,
                                  cells_use=NULL,
                                  slot="scale.data",
-                                 priority="avg_logFC",
+                                 priority="avg_log2FC",
                                  row_names_gp=10,
                                  sub_group=NULL,
                                  disp_min=-2.5,
@@ -574,10 +574,10 @@ markerComplexHeatmap <- function(seurat_object,
                "it is avaliable via devtools here: https://github.com/jokergoo/ComplexHeatmap"))
   }
 
-  if(priority=="avg_logFC") {
+  if(priority=="avg_log2FC") {
     top_markers <- marker_table %>%
     group_by(cluster) %>%
-    top_n(n=n_markers,wt=avg_logFC)
+    top_n(n=n_markers,wt=avg_log2FC)
   } else if (priority=="min_logFC") {
     top_markers <- marker_table %>%
       group_by(cluster) %>%