Overview
This package is a generalised container for a variety of front-end
function that execute complex workflows embedded in docker container.
The package includes a set of core functions which are required to allow
the front-end function to work in different environments like MAC OS,
Windows and Linux. The over all structure of the front-end is shown in
figure1
The package contains also a very simple skeleton function that can be
used as template for the creation of a new function. the elements to be
modified in the skeleton are shown in figure 2,
Panel A show the skeleton and panel B the modifications made in the
ribomito_qc function. IMPORTANT: The input-dir_path in
sketelon function must not be removed! It refer to the folder where the
data are located in your computer and this folder will be mount as
/scratch in the docker container. Here, instead you can find the header
of the R script present in the repbioinfo/ribomito docker container.
#!/usr/bin/env Rscript
library(data.table)
library(Matrix)
library(ggplot2)
# versione 3.1 mat_dgC <- as(mat_dgC, "dgCMatrix") was changed in mat_dgC <- as(mat_dgC, "CsparseMatrix"), since dgCMatrix is deprecated
# expression binning was inserted in the in the output file
...
#Version 3.6 support visium HD spatial transcriptomics output: embedding bc2cellid function, see below
setwd("/scratch")
args <- commandArgs(trailingOnly = TRUE)
if (length(args) < 1) {
stop("Usage: Rscript Ribomito_v3.6.2.R <infile> <rownames_format> <kit> [beads_location] [counts_threshold] [out_metrics] [out_freq] [mtx_file] [barcodes_file] [features_file]")
}
infile <- ifelse(length(args) >= 1, args[1])
rownames_format <- ifelse(length(args) >= 2, args[2], "symbol")
kit <- ifelse(length(args) >= 3, args[3], "scRNAseq")
beads_location <- if (length(args) >= 4) args[4] else NULL
counts_threshold <- ifelse(length(args) >= 5, as.numeric(args[5]), 2)
out_metrics <- ifelse(length(args) >= 6, args[6], "ribomito_metrics.csv")
out_freq <- ifelse(length(args) >= 7, args[7], "ribomito_freq.csv")
mtx_file <- if (length(args) >= 8) args[8] else NULL
barcodes_file <- if (length(args) >= 9) args[9] else NULL
features_file <- if (length(args) >= 10) args[10] else NULL
#' ribomito_qc
#'
#' Compute per-cell QC metrics: N.genes (counts > 2), ribo.score (%), mito.score (%)
#' Supports:
...
ribomito_qc <- function(infile,
rownames_format = c("ensemblID:symbol", "symbol"),
kit = c("scRNAseq", "scMultiomics10X", "Seeker", "visiumHD"),
beads_location = NULL,
counts_threshold = 2,
out_metrics = "ribomito_metrics.csv",
out_freq = "ribomito_Ngenes_freq.csv",
mtx_file = NULL,
barcodes_file = NULL,
features_file = NULL) {
...
}
ribomito_qc(infile=infile, rownames_format=rownames_format, kit=kit, beads_location=beads_location, out_metrics=out_metrics, out_freq=out_freq,
mtx_file=mtx_file, barcodes_file=barcodes_file, features_file=features_file, counts_threshold=counts_threshold)As you can see the parameters provided by the front-end function will be passed to this script and the script will take care of executing the code requested for the analysis.
Panel A reports the percentage of counts associated with mitochondrial
proteins plotted against the percentage associated with ribosomal
proteins. Each dot represents a cell and is colored according to the
number of detected genes. As above, we used the threshold defined in the
