kraken16S_container_noBLAST+.def

Bootstrap: docker
From: centos:latest

%environment
    export PATH=/opt/programs/ncbi-blast/bin:"$PATH"
    export PATH=/opt/programs/kraken2:"$PATH"
    export PATH=/opt/programs/BMTagger:"$PATH"
    export PATH=/opt/programs/Bracken:"$PATH"

%post
    cd /etc/yum.repos.d/
    sed -i 's/mirrorlist/#mirrorlist/g' /etc/yum.repos.d/CentOS-*
    sed -i 's|#baseurl=http://mirror.centos.org|baseurl=http://vault.centos.org|g' /etc/yum.repos.d/CentOS-*
    dnf install -y --nogpgcheck https://mirrors.rpmfusion.org/free/el/rpmfusion-free-release-$(rpm -E %rhel).noarch.rpm
    dnf update libmodulemd -y
    dnf install -y 'dnf-command(config-manager)'
    dnf config-manager --set-enabled powertools
    dnf upgrade --refresh -y
    dnf -y install util-linux-user autoconf automake R wget cmake git pigz gcc-c++ zlib-devel.x86_64 libstdc++-static libcurl-devel openssl-devel
    yum install -y python39-devel
    unlink /usr/bin/python3
    ln -s /usr/bin/python3.9 /usr/bin/python3
    ln -s /usr/bin/python3 /usr/bin/python
    ln -s /usr/bin/pip3 /usr/bin/pip
    pip3 install --upgrade pip
    pip3 install pandas numpy
    mkdir /opt/programs
    mkdir /mnt/databases
    #FASTP
    cd /usr/bin
    wget http://opengene.org/fastp/fastp
    chmod a+x ./fastp
    #BMTagger
    cd /opt/programs && mkdir BMTagger && cd BMTagger
    wget --reject fa,wbm ftp.ncbi.nlm.nih.gov:pub/agarwala/bmtagger/*
    chmod 755 *
    mkdir /mnt/databases/bmtagger
    #KneadData
    pip install kneaddata==0.7.2
    cd /usr/bin
    wget https://github.com/Benson-Genomics-Lab/TRF/releases/download/v4.09.1/trf409.linux64 -O trf
    chmod +x trf
    #Kraken2
    cd /opt/programs
    git clone https://github.com/DerrickWood/kraken2
    cd kraken2
    ./install_kraken2.sh .
    sed -i 's/FTP_SERVER="ftp/FTP_SERVER="https/' /opt/programs/kraken2/download_genomic_library.sh
    sed -i 's/s#^ftp/s#^https/' /opt/programs/kraken2/rsync_from_ncbi.pl
    mkdir /mnt/databases/kraken_db
    cd /mnt/databases/kraken_db/
    /opt/programs/kraken2/kraken2-build --db silvaNR99 --special silva --threads 24
    cd silvaNR99
    wget https://www.arb-silva.de/fileadmin/silva_databases/release_138_1/Exports/taxonomy/tax_slv_ssu_138.1.tre.gz
    pigz -d tax_slv_ssu_138.1.tre.gz
    #Bracken
    cd /opt/programs
    git clone https://github.com/jenniferlu717/Bracken.git
    cd Bracken
    bash install_bracken.sh
    cd /mnt/databases/kraken_db
    export PATH=/opt/programs/kraken2:"$PATH"
    /opt/programs/Bracken/bracken-build -d silvaNR99 -t 24 -k 35 -l 250
    rm -fr silvaNR99/database.kraken
    /opt/programs/Bracken/bracken-build -d silvaNR99 -t 24 -k 35 -l 500
    #Otu table custom processing scripts
    cd /usr/bin
    wget https://raw.githubusercontent.com/FabbriniMarco/16S-KrakenBracken-Pipeline/main/CPU_process_otutab.py
    wget https://raw.githubusercontent.com/FabbriniMarco/16S-KrakenBracken-Pipeline/main/CPU_process_otutab_relabb.py
    wget https://raw.githubusercontent.com/FabbriniMarco/16S-KrakenBracken-Pipeline/main/compute_diversity_kraken.r
    chmod +x CPU_process_otutab*
    chmod +x compute_diversity_kraken.r
    #Create workdir
    cd
    mkdir /workdir
    chmod 777 /workdir
    #Setup required R packages
    echo -e ' 
    install.packages("ape", repos="https://cran.mirror.garr.it/CRAN/")
    install.packages("BiocManager", repos="https://cran.mirror.garr.it/CRAN/")
    BiocManager::install("phyloseq", update=FALSE)
    install.packages("parallel", repos="https://cran.mirror.garr.it/CRAN/")
    install.packages("doParallel", repos="https://cran.mirror.garr.it/CRAN/")
    install.packages("vegan", repos="https://cran.mirror.garr.it/CRAN/")
    install.packages("otuSummary", repos="https://cran.mirror.garr.it/CRAN/")
    install.packages("picante", repos="https://cran.mirror.garr.it/CRAN/")
    install.packages("openxlsx", repos="https://cran.mirror.garr.it/CRAN/") ' > r_install_packages.R
    Rscript r_install_packages.R
    rm -fr r_install_packages.R
    #Get the pipeline script
    cd /usr/bin
    wget https://raw.githubusercontent.com/FabbriniMarco/16S-KrakenBracken-Pipeline/main/kraken16S.sh -O kraken16S
    chmod +x kraken16S

%labels
    Author Marco Fabbrini - fabbrinimarco.mf@gmail.com
    Version v1