Install and run R package mappable on SLURM HPC cluster

mappable.sh

#!/bin/bash
#SBATCH --nodes 1
#SBATCH --partition phi #general #general_requeue
#SBATCH --ntasks 1 #4 #24
#-SBATCH --cpus-per-task 64
#SBATCH --output mappable.out
#SBATCH --job-name kmap

genomes='genomes <- c(
    "BSgenome.Dmelanogaster.UCSC.dm6",
    "BSgenome.Hsapiens.UCSC.hg19",
    "BSgenome.Hsapiens.UCSC.hg38",
    "BSgenome.Hsapiens.NCBI.GRCh38"
)'

main() {
    local bootstrap
    bootstrap=

    [[ -n ${SLURM_JOB_ID} ]] && echo ${SLURM_JOB_ID} >> mappable.out
    module purge
    if [[ -n "$bootstrap" ]]; then
	install_and_r_and_pkgs
    fi
    module load git
    update_kmap
    run_kmap
}

install_and_r_and_pkgs() {
    conda install -y -c r r-essentials r-xml r-devtools
    Rscript -e 'source("https://bioconductor.org/biocLite.R")'
}

# Globals: genomes
update_kmap() {
    R -q --vanilla <<EOF
$genomes
install_genome <- function(genome) {
    if (! requireNamespace(genome, quietly = TRUE)) {
        BiocInstaller::biocLite(genome)
    }
}
invisible(lapply(genomes, install_genome))

devtools::install_github("coregenomics/kmap",
                         repos = BiocInstaller::biocinstallRepos())
EOF
}

run_kmap() {
    R -q --vanilla <<EOF
$genomes
dump_and_quit <- function() {
    # Save debugging info to file last.dump.rda
    dump.frames(to.file = TRUE)
    # Quit R with error status
    q(status = 1)
}
options(error = dump_and_quit)
library(kmap)
invisible(lapply(genomes, mappable))

message("Completed all genomes!")
EOF
}

main