#!/usr/bin/env Rscript # ============================================================================= # BioF3 one-shot data & package bootstrap # ============================================================================= # # Installs every R package referenced by the BioF3 companion scripts, then # downloads the three public 10x datasets that cannot be shipped with a # Bioconductor package into ~/biof3-data/ (or $BIOF3_DATA_DIR). # # Every step is idempotent: # - installed packages are skipped # - already-downloaded data are not re-fetched # # Usage: # Rscript scripts/biof3_prepare_data.R # # Optional env vars: # BIOF3_DATA_DIR=/path/to/data default: ~/biof3-data # # ============================================================================= options(stringsAsFactors = FALSE, timeout = 900) # ---- logging helpers ---------------------------------------------------- .print_banner <- function(text) { line <- paste(rep("=", 70), collapse = "") cat("\n", line, "\n", text, "\n", line, "\n", sep = "") } .step <- function(text) cat("\n>>> ", text, "\n", sep = "") .ok <- function(text) cat(" [ok] ", text, "\n", sep = "") .warn <- function(text) cat(" [!!] ", text, "\n", sep = "") # ---- R version check ---------------------------------------------------- .print_banner("BioF3 data & package bootstrap") .step(paste0("R version: ", getRversion())) if (getRversion() < "4.2") { stop("R >= 4.2 is recommended. Please upgrade.") } # ---- install helpers ---------------------------------------------------- install_if_missing <- function(pkg, source = c("cran", "bioc", "github"), github_repo = NULL) { source <- match.arg(source) if (requireNamespace(pkg, quietly = TRUE)) { .ok(paste0(pkg, " already installed")) return(TRUE) } .step(paste0("installing ", pkg, " (", source, ")")) ok <- tryCatch({ if (source == "cran") { install.packages(pkg) } else if (source == "bioc") { if (!requireNamespace("BiocManager", quietly = TRUE)) { install.packages("BiocManager") } BiocManager::install(pkg, update = FALSE, ask = FALSE) } else if (source == "github") { if (!requireNamespace("devtools", quietly = TRUE)) { install.packages("devtools") } devtools::install_github(github_repo, upgrade = "never") } requireNamespace(pkg, quietly = TRUE) }, error = function(e) { .warn(paste0(pkg, " failed via primary source: ", conditionMessage(e))) FALSE }) if (!ok) { # Bioconductor fallback with explicit repos (works when BiocManager # version validation keeps failing). if (source == "bioc") { .step(paste0("retrying ", pkg, " via explicit Bioconductor mirrors")) repos <- c( BioCsoft = "https://bioconductor.org/packages/3.22/bioc", BioCann = "https://bioconductor.org/packages/3.22/data/annotation", BioCexp = "https://bioconductor.org/packages/3.22/data/experiment", CRAN = "https://mirrors.tuna.tsinghua.edu.cn/CRAN" ) ok <- tryCatch({ install.packages(pkg, repos = repos, dependencies = TRUE) requireNamespace(pkg, quietly = TRUE) }, error = function(e) { .warn(paste0(pkg, " failed on fallback: ", conditionMessage(e))) FALSE }) } } if (ok) .ok(paste0(pkg, " ok")) else .warn(paste0(pkg, " installation failed — please retry manually")) invisible(ok) } # ---- CRAN packages ------------------------------------------------------ .print_banner("Installing CRAN packages") cran_pkgs <- c( "Matrix", "dplyr", "tidyr", "readr", "stringr", "tibble", "ggplot2", "ggrepel", "patchwork", "pheatmap", "RColorBrewer", "viridis", "BiocManager", "devtools", "remotes", "Seurat", "SeuratObject", "sna", "ggnetwork", "collapse", # CellChat runtime deps "hash" # scRepertoire runtime dep ) for (p in cran_pkgs) install_if_missing(p, source = "cran") # ---- Bioconductor packages ---------------------------------------------- .print_banner("Installing Bioconductor packages") bioc_pkgs <- c( # single-cell "SingleCellExperiment", "SummarizedExperiment", "GenomicRanges", "slingshot", "tradeSeq", "scRepertoire", "Signac", "EnsDb.Hsapiens.v86", # bulk RNA-seq "DESeq2", "edgeR", "limma", "sva", "clusterProfiler", "org.Hs.eg.db", "AnnotationDbi", "enrichplot", "DOSE", "fgsea", "airway", "fission", "bladderbatch", "pasilla", "apeglm", "RUVSeq" ) for (p in bioc_pkgs) install_if_missing(p, source = "bioc") # ---- GitHub-only packages ----------------------------------------------- .print_banner("Installing GitHub-only packages") install_if_missing("CellChat", source = "github", github_repo = "jinworks/CellChat") install_if_missing("SeuratData", source = "github", github_repo = "satijalab/seurat-data") # SeuratData companion data package tryCatch({ if (requireNamespace("SeuratData", quietly = TRUE)) { installed <- SeuratData::InstalledData() if (!"stxBrain" %in% installed$Dataset) { .step("installing SeuratData::stxBrain (~136 MB)") SeuratData::InstallData("stxBrain") .ok("stxBrain installed") } else { .ok("stxBrain already installed") } } }, error = function(e) { .warn(paste0("stxBrain install failed: ", conditionMessage(e))) }) # ---- data directory layout ---------------------------------------------- .print_banner("Setting up data directory") data_root <- Sys.getenv("BIOF3_DATA_DIR", file.path(path.expand("~"), "biof3-data")) dir.create(data_root, recursive = TRUE, showWarnings = FALSE) .ok(paste0("data root: ", normalizePath(data_root))) # ---- dataset catalog ---------------------------------------------------- datasets <- list( pbmc3k = list( subdir = "pbmc3k", tarball = "pbmc3k.tar.gz", url = paste0( "https://cf.10xgenomics.com/samples/cell/pbmc3k/", "pbmc3k_filtered_gene_bc_matrices.tar.gz" ), size_mb = 7, sentinel = "filtered_gene_bc_matrices/hg19/matrix.mtx" ), pbmc5k_cite = list( subdir = "pbmc5k-citeseq", tarball = "5k_pbmc_filtered.tar.gz", url = paste0( "https://cf.10xgenomics.com/samples/cell-exp/3.1.0/", "5k_pbmc_protein_v3_nextgem/", "5k_pbmc_protein_v3_nextgem_filtered_feature_bc_matrix.tar.gz" ), size_mb = 37, sentinel = "filtered_feature_bc_matrix/matrix.mtx.gz" ), pbmc10k_atac = list( subdir = "pbmc10k-scatac", tarball = NULL, # these are single files, not tarballs files = list( h5 = list( name = "filtered_peak_bc_matrix.h5", url = paste0( "https://cf.10xgenomics.com/samples/cell-atac/2.1.0/", "10k_pbmc_ATACv2_nextgem_Chromium_Controller/", "10k_pbmc_ATACv2_nextgem_Chromium_Controller_", "filtered_peak_bc_matrix.h5" ) ), meta = list( name = "singlecell.csv", url = paste0( "https://cf.10xgenomics.com/samples/cell-atac/2.1.0/", "10k_pbmc_ATACv2_nextgem_Chromium_Controller/", "10k_pbmc_ATACv2_nextgem_Chromium_Controller_", "singlecell.csv" ) ) ), size_mb = 200, sentinel = "filtered_peak_bc_matrix.h5" ) ) download_dataset <- function(name, info) { .step(paste0("preparing ", name, " (~", info$size_mb, " MB)")) target_dir <- file.path(data_root, info$subdir) dir.create(target_dir, recursive = TRUE, showWarnings = FALSE) # already ready? sentinel_path <- file.path(target_dir, info$sentinel) if (file.exists(sentinel_path)) { .ok(paste0(name, " already present")) return(TRUE) } if (!is.null(info$tarball)) { tar_path <- file.path(target_dir, info$tarball) if (!file.exists(tar_path)) { .step(paste0("downloading ", basename(info$url))) tryCatch( download.file(info$url, destfile = tar_path, mode = "wb"), error = function(e) { .warn(paste0("download failed: ", conditionMessage(e))) } ) } if (file.exists(tar_path)) { .step("extracting tarball") utils::untar(tar_path, exdir = target_dir) } } else if (!is.null(info$files)) { for (f in info$files) { file_path <- file.path(target_dir, f$name) if (!file.exists(file_path)) { .step(paste0("downloading ", f$name)) tryCatch( download.file(f$url, destfile = file_path, mode = "wb"), error = function(e) { .warn(paste0("download failed: ", conditionMessage(e))) } ) } } } if (file.exists(sentinel_path)) { .ok(paste0(name, " ready")) TRUE } else { .warn(paste0(name, " did not complete (check network)")) FALSE } } .print_banner("Downloading 10x Genomics public datasets") for (n in names(datasets)) { download_dataset(n, datasets[[n]]) } # ---- summary ------------------------------------------------------------ .print_banner("Summary") cat("Installed packages (examples):\n") samples <- c("Seurat", "DESeq2", "edgeR", "clusterProfiler", "Signac", "scRepertoire", "airway", "fission") for (p in samples) { flag <- if (requireNamespace(p, quietly = TRUE)) "ok" else "MISSING" cat(sprintf(" %-20s [%s]\n", p, flag)) } cat("\nData directory:", data_root, "\n") for (n in names(datasets)) { info <- datasets[[n]] sentinel <- file.path(data_root, info$subdir, info$sentinel) flag <- if (file.exists(sentinel)) "ok" else "MISSING" cat(sprintf(" %-20s [%s]\n", info$subdir, flag)) } cat("\nDone. Any items flagged MISSING can be retried by rerunning this script.\n")