@sjcrh/proteinpaint-server 2.122.0 → 2.123.0

package/utils/edge.R

@@ -1,283 +0,0 @@
-# Test syntax: cat ~/sjpp/test.txt | time Rscript edge.R
-
-# Load required packages
-suppressWarnings({
-  library(jsonlite)
-  library(rhdf5)
-  library(stringr)
-  library(readr)
-  suppressPackageStartupMessages(library(edgeR))
-  suppressPackageStartupMessages(library(dplyr))
-})
-
-# Filter based on CPM
-filter_using_cpm <- function(y, gene_cpm_cutoff, sample_cpm_cutoff, count_cpm_cutoff) {
-  selr <- rowSums(cpm(y$counts)>gene_cpm_cutoff)>=sample_cpm_cutoff
-  selc <- colSums(cpm(y$counts))>=count_cpm_cutoff
-  y <- y[selr, selc]
-}
-
-# Read JSON input from stdin
-read_json_time <- system.time({
-  con <- file("stdin", "r")
-  json <- readLines(con, warn=FALSE)
-  close(con)
-  input <- fromJSON(json)
-  cases <- unlist(strsplit(input$case, ","))
-  controls <- unlist(strsplit(input$control, ","))
-  combined <- c("geneID", "geneSymbol", cases, controls)
-})
-#cat("Time to read JSON: ", as.difftime(read_json_time, units = "secs")[3], " seconds\n")
-
-# Read counts data
-read_counts_time <- system.time({
-  if (input$storage_type == "HDF5") {
-    geneIDs <- h5read(input$input_file, "gene_ids")
-    geneNames <- h5read(input$input_file, "gene_names")
-    samples <- h5read(input$input_file, "samples")
-
-    # Find indices of case and control samples in the HDF5 file
-    case_indices <- match(cases, samples)
-    control_indices <- match(controls, samples)
-
-    # Check for missing samples
-    if (any(is.na(case_indices))) {
-      missing_cases <- cases[is.na(case_indices)]
-      stop(paste(missing_cases, "not found"))
-    }
-    if (any(is.na(control_indices))) {
-      missing_controls <- controls[is.na(control_indices)]
-      stop(paste(missing_controls, "not found"))
-    }
-
-    samples_indices <- c(case_indices, control_indices)
-    read_counts <- as.data.frame(t(h5read(input$input_file, "counts", index = list(samples_indices, 1:length(geneIDs)))))
-    colnames(read_counts) <- c(cases, controls)
-  } else if (input$storage_type == "text") {
-    suppressWarnings({
-      suppressMessages({
-        read_counts <- read_tsv(input$input_file, col_names = TRUE, col_select = combined)
-      })
-    })
-    geneIDs <- unlist(read_counts[1])
-    geneNames <- unlist(read_counts[2])
-    read_counts <- select(read_counts, -geneID)
-    read_counts <- select(read_counts, -geneSymbol)
-  } else {
-    stop("Unknown storage type")
-  }
-})
-#cat("Time to read counts data: ", as.difftime(read_counts_time, units = "secs")[3], " seconds\n")
-
-# Create conditions vector
-conditions <- c(rep("Diseased", length(cases)), rep("Control", length(controls)))
-gene_id_symbols <- paste0(geneIDs, "\t", geneNames)
-
-# Create DGEList object
-dge_list_time <- system.time({
-  y <- DGEList(counts = read_counts, group = conditions, genes = gene_id_symbols)
-})
-#cat("Time to generate DGEList: ", as.difftime(dge_list_time, units = "secs")[3], " seconds\n")
-
-# Filter and normalize counts
-filter_time <- system.time({
-  keep <- filterByExpr(y, min.count = input$min_count, min.total.count = input$min_total_count)
-})
-#cat("Time to filter by expression: ", as.difftime(filter_time, unit = "secs")[3], " seconds\n")
-
-normalization_time <- system.time({
-  y <- y[keep, keep.lib.sizes = FALSE]
-  y <- normLibSizes(y) # Using TMM method for normalization
-})
-#cat("Normalization time: ", as.difftime(normalization_time, units = "secs")[3], " seconds\n")
-
-# Cutoffs for cpm, will add them as UI options later
-if (length(samples_indices) > 100) {
-  gene_cpm_cutoff <- 15
-  sample_cpm_cutoff <- 30
-  count_cpm_cutoff <- 100000
-} else {
-  gene_cpm_cutoff <- 5
-  sample_cpm_cutoff <- 15
-  count_cpm_cutoff <- 100000
-}
-
-filter_using_cpm_time <- system.time({
-  y <- filter_using_cpm(y, gene_cpm_cutoff, sample_cpm_cutoff, count_cpm_cutoff) # Filtering counts matrix based on gene_cpm_cutoff, sample_cpm_cutoff and count_cpm_cutoff
-})
-#cat("Filter using cpm time: ", as.difftime(filter_using_cpm_time, units = "secs")[3], " seconds\n")
-
-if (dim(y)[1]==0) { # Its possible after filtering there might not be any genes left in the matrix, in such a case the R code must exit gracefully with an error.
-  stop("Number of genes after filtering = 0, cannot proceed any further") 
-}  
-if (dim(y)[2]==0) { # Its possible after filtering there might not be any samples left in the matrix, in such a case the R code must exit gracefully with an error.
-  stop("Number of samples after filtering = 0, cannot proceed any further") 
-}
-
-# Saving MDS plot image
-
-if (dim(read_counts)[1] * dim(read_counts)[2] < as.numeric(input$mds_cutoff)) { # If the dimensions of the read counts matrix is below this threshold, only then the mds image will be generated as its very compute intensive
-  mds_plot_time <- system.time({
-    set.seed(as.integer(Sys.time())) # Set the seed according to current time
-    cachedir <- input$cachedir # Importing serverconfig.cachedir
-    random_number <- runif(1, min = 0, max = 1) # Generating random number
-    mds_image_name <- paste0("edgeR_mds_temp_",random_number,".png") # Generating random image name so that simultaneous server side requests do NOT generate the same edgeR file name
-    png(filename = paste0(cachedir,"/",mds_image_name), width = 1000, height = 1000, res = 200) # Opening a png device
-    par(oma = c(0, 0, 0, 0)) # Creating a margin
-    mds_conditions <- c(rep("T", length(cases)), rep("C", length(controls))) # Case samples are labelled "T" and control samples are labelled "C". Single-letter labelling added because otherwise labels get overwritten on each other.
-    plotMDS(y, labels = mds_conditions) # Plot the edgeR MDS plot
-    # dev.off() # Gives a null device message which breaks JSON. Commenting it out for now, will investigate it later
-  })
-  #cat("mds plot time: ", as.difftime(mds_plot_time, units = "secs")[3], " seconds\n")
-}
-
-# Differential expression analysis
-if (length(input$conf1) == 0) { # No adjustment of confounding factors
-  design <- model.matrix(~conditions) # Based on the protocol defined in section 1.4 of edgeR manual https://bioconductor.org/packages/release/bioc/vignettes/edgeR/inst/doc/edgeRUsersGuide.pdf
-} else { # Adjusting for confounding factors
-  # Check the type of confounding variable
-  if (input$conf1_mode == "continuous") { # If this is float, the input conf1 vector should be converted into a numeric vector
-    conf1 <- as.numeric(input$conf1)
-  } else { # When input$conf1_mode == "discrete" keep the vector as string.
-    conf1 <- as.factor(input$conf1)
-  }
-
-  if (length(input$conf2) == 0) { # No adjustment of confounding factor 2
-    y$samples <- data.frame(y$samples, conditions = conditions, conf1 = conf1)
-    model_gen_time <- system.time({
-      design <- model.matrix(~ conditions + conf1, data = y$samples)
-    })
-    #cat("Time for making design matrix: ", as.difftime(model_gen_time, units = "secs")[3], " seconds\n")
-  } else {
-    # Check the type of confounding variable 2
-    if (input$conf2_mode == "continuous") { # If this is float, the input conf2 vector should be converted into a numeric vector
-      conf2 <- as.numeric(input$conf2)
-    } else { # When input$conf2_mode == "discrete" keep the vector as string.
-      conf2 <- as.factor(input$conf2)
-    }
-    y$samples <- data.frame(y$samples, conditions = conditions, conf1 = conf1, conf2 = conf2)
-    model_gen_time <- system.time({
-      design <- model.matrix(~ conditions + conf1 + conf2, data = y$samples)
-    })
-    #cat("Time for making design matrix: ", as.difftime(model_gen_time, units = "secs")[3], " seconds\n")
-
-  }
-}
-
-DE_method <- input$DE_method
-if (DE_method == "edgeR") {
-  fit_time <- system.time({
-    suppressWarnings({
-      suppressMessages({
-        fit <- glmQLFit(y,design) # The glmQLFit() replaces glmFit() which implements the quasi-likelihood function. This is better able to account for overdispersion as it employs a more lenient approach where variance is not a fixed function of the mean.
-      })
-    })
-  })
-  #cat("QL fit time: ", as.difftime(fit_time, units = "secs")[3], " seconds\n")
-  test_time <- system.time({
-    suppressWarnings({
-      suppressMessages({
-        et <- glmQLFTest(fit, coef = "conditionsDiseased")
-      })
-    })
-  })
-  #cat("QL test time: ", as.difftime(test_time, units = "secs")[3], " seconds\n")    
-  
-  # Saving QL fit image
-  ql_plot_time <- system.time({
-    set.seed(as.integer(Sys.time())) # Set the seed according to current time
-    cachedir <- input$cachedir # Importing serverconfig.cachedir
-    random_number <- runif(1, min = 0, max = 1) # Generating random number
-    fit_image_name <- paste0("edgeR_ql_temp_",random_number,".png") # Generating random image name so that simultaneous server side requests do NOT generate the same edgeR file name
-    png(filename = paste0(cachedir,"/",fit_image_name), width = 1000, height = 1000, res = 200) # Opening a png device
-    par(oma = c(0, 0, 0, 0)) # Creating a margin
-    plotQLDisp(fit) # Plot the edgeR fit
-    # dev.off() # Gives a null device message which breaks JSON. Commenting it out for now, will investigate it later
-  })
-  #cat("ql plot time: ", as.difftime(ql_plot_time, units = "secs")[3], " seconds\n")
-  logfc <- et$table$logFC
-  logcpm <- et$table$logCPM
-  pvalues <- et$table$PValue
-  genes_matrix <- str_split_fixed(unlist(et$genes), "\t", 2)
-  geneids <- unlist(genes_matrix[, 1])
-  geneNames <- unlist(genes_matrix[, 2])
-  adjust_p_values <- p.adjust(pvalues, method = "fdr")
-} else if (DE_method == "limma") {
-  # Do voom transformation and fit linear model
-  voom_transformation_lmfit_time <- system.time({
-    suppressWarnings({
-      suppressMessages({
-        set.seed(as.integer(Sys.time())) # Set the seed according to current time
-        cachedir <- input$cachedir # Importing serverconfig.cachedir
-        random_number <- runif(1, min = 0, max = 1) # Generating random number
-        fit_image_name <- paste0("limma_voom_temp_",random_number,".png") # Generating random image name so that simultaneous server side requests do NOT generate the same edgeR file name
-        png(filename = paste0(cachedir,"/",fit_image_name), width = 1000, height = 1000, res = 200) # Opening a png device
-        par(oma = c(0, 0, 0, 0)) # Creating a margin
-        suppressWarnings({
-          suppressMessages({
-            fit <- voomLmFit(y, design, plot = TRUE) # This is base don the recommendation of the edgeR limma/voom authors https://support.bioconductor.org/p/9161585/
-          })
-        })
-        dev.off() # Gives a null device message which breaks JSON. Commenting it out for now, will investigate it later
-      })
-    })
-  })
-  #cat("voom transformation + limma fit time: ", as.difftime(voom_transformation_lmfit_time, units = "secs")[3], " seconds\n")
-
-  # Saving mean-difference plot (aka MA plot)
-  #set.seed(as.integer(Sys.time())) # Set the seed according to current time
-  #cachedir <- input$cachedir # Importing serverconfig.cachedir
-  #random_number <- runif(1, min = 0, max = 1) # Generating random number
-  #md_image_name <- paste0("limma_md_temp_",random_number,".png") # Generating random image name so that simultaneous server side requests do NOT generate the same edgeR file name
-  #png(filename = paste0(cachedir,"/",md_image_name), width = 1000, height = 1000, res = 200) # Opening a png device
-  #par(oma = c(0, 0, 0, 0)) # Creating a margin
-  #plotMD(fit) # Plot the limma fit
-  ## dev.off() # Gives a null device message which breaks JSON. Commenting it out for now, will investigate it later
-
-  # Empirical Bayes smoothing
-  empirical_smoothing_time <- system.time({
-    suppressWarnings({
-      suppressMessages({
-        tmp <- eBayes(fit)
-      })
-    })
-  })
-  #cat("Empirical smoothing time: ", as.difftime(empirical_smoothing_time, units = "secs")[3], " seconds\n")
-
-  # Time for selecting top genes
-  top_genes_selection_time <- system.time({
-    suppressWarnings({
-      suppressMessages({
-        top_table <- topTable(tmp, coef = "conditionsDiseased", number = Inf, adjust.method = "fdr") # The coeff needs to be specified in topTable() because it needs to know for which contrast the logFC needs to be calculated https://www.biostars.org/p/160465/
-        logfc <- top_table$logFC
-        pvalues <- top_table$P.Value
-        genes_matrix <- str_split_fixed(unlist(top_table$genes), "\t", 2)
-        geneids <- unlist(genes_matrix[, 1])
-        geneNames <- unlist(genes_matrix[, 2])
-        adjust_p_values <- top_table$adj.P.Val
-      })
-    })
-  })
-  #cat("Time for selecting top genes: ", as.difftime(top_genes_selection_time, units = "secs")[3], " seconds\n")
-} else { # Should not happen
-  stop(paste0("Unknown method:", DE_method))
-}
-
-final_data_generation_time <- system.time({
-  output <- data.frame(geneids, geneNames, logfc, pvalues, adjust_p_values)
-  names(output)[1] <- "gene_id"
-  names(output)[2] <- "gene_name"
-  names(output)[3] <- "fold_change"
-  names(output)[4] <- "original_p_value"
-  names(output)[5] <- "adjusted_p_value"
-})
-final_output <- c()
-final_output$gene_data <- output
-final_output$edgeR_ql_image_name <- fit_image_name
-if (dim(read_counts)[1] * dim(read_counts)[2] < as.numeric(input$mds_cutoff)) { # If the dimensions of the read counts matrix is below this threshold, only then the mds image will be generated as its very compute intensive
-  final_output$edgeR_mds_image_name <- mds_image_name
-}
-#cat("Time for generating final dataframe: ", as.difftime(final_data_generation_time, unit = "secs")[3], " seconds\n")
-
-# Output results
-toJSON(final_output, digits = NA, na = "string") # Setting digits = NA makes toJSON() use the max precision. na='string' causes any "not a number" to be reported as string. This from ?toJSON() documentation

package/utils/fdr.R

@@ -1,9 +0,0 @@
-argv <- commandArgs(TRUE)
-
-infile <- argv[1]
-outfile <- argv[2]
-
-dat <- read.table(infile,sep="\t",header=F,quote="")
-out <- p.adjust( dat, method="BH")
-
-write.table(out,file=outfile,sep="\t",quote=F,row.names=F,col.names=F)

package/utils/fisher.2x3.R

@@ -1,12 +0,0 @@
-out <- NULL
-con <- file("stdin","r")
-dat <- read.table(con,sep="\t",header=F,quote="")
-
-for (i in 1:nrow(dat)) {
-	x <- fisher.test( matrix( c( dat[i,2], dat[i,3], dat[i,4], dat[i,5], dat[i,6], dat[i,7] ), nrow=2 ) )
-	out <- rbind(out,x$p.value)
-}
-out <- cbind(dat,out)
-
-write.table(out,file="",sep="\t",quote=F,row.names=F,col.names=F)
-close(con)

package/utils/fisher.R

@@ -1,9 +0,0 @@
-con <- file("stdin","r")
-dat <- read.table(con,sep="\t",header=F,quote="")
-
-pvals <- apply(dat[,2:5], 1, function(row) fisher.test(matrix(row, ncol=2))$p.value)
-
-out <- cbind(dat,pvals)
-
-write.table(out,file="",sep="\t",quote=F,row.names=F,col.names=F)
-close(con)