#!/bin/bash

###I###R###D######U###2###3###3#######T###R###A###N###S###V###I###H###M###I####
###                                                                         ###
###           /\  ______      ____    _____     ___     ______  /\          ###
###           ||  \ \ \ \    (  _ \  (  _  )   / __)   / / / /  ||          ###
###           ||   > > > >    )   /   )(_)(   ( (__   < < < <   ||          ###
###           ||  /_/_/_/    (_)\_)  (___/\\   \___)   \_\_\_\  ||          ###
###           \/                                                \/          ###
###                                                                         ###
###I###R###D######U###2###3###3#######T###R###A###N###S###V###I###H###M###I####
# Name ___________________ Start_RQC.sh
# Version ________________ v.2023.06
# Author _________________ Nicolas Fernandez
# Affiliation ____________ IRD_U233_TransVIHMI
# Aim ____________________ Bash script running quality_control.smk snakefile
# Date ___________________ 2021.10.12
# Latest modifications ___ 2023.06.21
# Use ____________________ bash Start_RQC.sh

###############################################################################
### COLORS ###
##############
red="\033[1;31m"   # red
green="\033[1;32m" # green
ylo="\033[1;33m"   # yellow
blue="\033[1;34m"  # blue
nc="\033[0m"       # no color

###############################################################################
### ABOUT ###
#############
workdir=$(cd "$(dirname "${BASH_SOURCE[0]}" )" && pwd) # Get working directory
rqc_version="2023.06"                                  # RQC version
workflow_base_version="2023.06"                        # Workflow base version

echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}ABOUT${nc} ${green}#####${nc}
${green}-----------------${nc}

${blue}Name${nc} ___________________ Start_RQC.sh
${blue}Version${nc} ________________ ${ylo}${rqc_version}${nc}
${blue}Author${nc} _________________ Nicolas Fernandez
${blue}Affiliation${nc} ____________ IRD_U233_TransVIHMI
${blue}Aim${nc} ____________________ Bash script for ${red}R${nc}reads ${red}Q${nc}uality ${red}C${nc}ontrol
${blue}Date${nc} ___________________ 2021.10.12
${blue}Latest modifications${nc} ___ 2023.06.21
${blue}Run${nc} ____________________ bash Start_RQC.sh
"

###############################################################################
### OPERATING SYSTEM ###
########################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}OPERATING SYSTEM${nc} ${green}#####${nc}
${green}----------------------------${nc}
"

# Get and print operating system 
case "$OSTYPE" in
  linux*)   os="linux" ;;
  bsd*)     os="bsd" ;;
  darwin*)  os="osx" ;;
  solaris*) os="solaris" ;;
  msys*)    os="windows" ;;
  cygwin*)  os="windows" ;;
  *)        os="unknown (${OSTYPE})" ;;
esac
echo -e "${blue}Operating system${nc} _______ ${red}${os}${nc}"

# Get and print shell
shell=$SHELL
echo -e "${blue}Shell${nc} __________________ ${ylo}${shell}${nc}"

###############################################################################
### HARDWARE ###
################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}HARDWARE${nc} ${green}#####${nc}
${green}--------------------${nc}
"

if [[ ${os} == "osx" ]]
then
    model_name=$(sysctl -n machdep.cpu.brand_string) # Get chip model name
    physical_cpu=$(sysctl -n hw.physicalcpu)         # Get physical cpu
    logical_cpu=$(sysctl -n hw.logicalcpu)           # Get logical cpu
    mem_size=$(sysctl -n hw.memsize)                 # Get memory size (bit)
    ram_gb=$(expr ${mem_size} \/ $((1024**3)) )      # mem_size / 1024**3 = Gb
elif [[ ${os} == "linux" ]]
then
    model_name=$(lscpu | grep -o -E "Model name: +.+" | sed -E "s/Model name: +//")                           # Get chip model name
    physical_cpu=$(lscpu | grep -o -E "^CPU\(s\): +[0-9]+" | sed -E "s/CPU\(s\): +//")                        # Get physical cpu
    threads_cpu=$(lscpu | grep -o -E "^Thread\(s\) per core: +[0-9]+" | sed -E "s/Thread\(s\) per core: +//") # Get thread(s) per core
    logical_cpu=$(expr ${physical_cpu} \* ${threads_cpu})                                                     # Calcul logical cpu
    mem_size=$(grep -o -E "MemTotal: +[0-9]+" /proc/meminfo | sed -E "s/MemTotal: +//")                       # Get memory size (Kb)
    ram_gb=$(expr ${mem_size} \/ $((1024**2)) )                                                               # mem_size / 1024**2 = Gb
else
    echo -e "Please, use '${red}osx${nc}' or '${red}linux${nc}' operating systems"
    exit 1
fi

# Print some hardware specifications (maybe wrong with WSL...)
echo -e "                         ${ylo}Brand(R)${nc} | ${ylo}Type(R)${nc} | ${ylo}Model${nc} | ${ylo}@ Speed GHz${nc}
${blue}Chip Model Name${nc} ________ ${model_name}
${blue}Physical CPUs${nc} __________ ${red}${physical_cpu}${nc}
${blue}Logical CPUs${nc} ___________ ${red}${logical_cpu}${nc} threads
${blue}System Memory${nc} __________ ${red}${ram_gb}${nc} Gb of RAM
"

###############################################################################
### WORKFLOW-BASE INSTALLATION ###
##################################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}WORKFLOW-BASE INSTALLATION${nc} ${green}#####${nc}
${green}---------------------------------------${nc}
"

# Test if latest 'workflow-base' environment exist
if [[ $(conda info --envs | grep -o -E "^workflow-base_v.${workflow_base_version}") ]]
then
    echo -e "
Conda environment ${ylo}workflow-base_v.${workflow_base_version}${nc} it's already created!
"
else
    echo -e "
Conda environment ${red}workflow-base_v.${workflow_base_version}${nc} will be now created, with:

    # ${red}Snakemake${nc}: Run RQC workflow (ver. 7.28.3)
    # ${red}Mamba${nc}:     Install snakemake conda's environments, faster than conda (ver. 1.4.4)
    # ${red}Yq${nc}:        Parse config.yaml file (ver. 3.2.2)
    # ${red}Rename${nc}:    Rename fastq files (ver. 1.601)
    # ${red}Graphviz${nc}:  Dot snakemake DAG (ver. 8.0.5)
"
    conda env create -f ${workdir}/workflow/environments/${os}/workflow-base_v.${workflow_base_version}.yaml
fi

# Remove depreciated 'gevarli' or 'snakemake' old environments
old_envs="gevarli-base_v.2022.11 \
          gevarli-base_v.2022.12 \
          gevarli-base_v.2023.01 \
          gevarli-base_v.2023.02 \
          gevarli-base_v.2023.03 \
          gevarli-base_v.2023.04 \
          snakemake-base_v.2023.01 \
          snakemake-base_v.2023.02 \
          snakemake-base_v.2023.03 \
          snakemake-base_v.2023.04"

for env in ${old_envs} ; do
    conda remove --name ${env} --all --yes --quiet 2> /dev/null ;
done

###############################################################################
### CONDA ACTIVATION ###
########################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}CONDA ACTIVATION${nc} ${green}#####${nc}
${green}----------------------------${nc}
"

echo -e "conda activate ${ylo}workflow-base_v.${workflow_base_version}${nc}"

# intern shell source conda
source ~/miniconda3/etc/profile.d/conda.sh 2> /dev/null          # local user
source /usr/local/miniconda3/etc/profile.d/conda.sh 2> /dev/null # HPC server
conda activate workflow-base_v.${workflow_base_version}          # conda activate workflow-base

###############################################################################
### SETTINGS ###
################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}SETTINGS${nc} ${green}#####${nc}
${green}--------------------${nc}
"

conda_version=$(conda --version | sed 's/conda //')                   # Conda version (ver. >= 23.3.1)
snakemake_version=$(snakemake --version)                              # Snakemake version (ver. 7.25.0)
mamba_version=$(mamba --version | sed 's/mamba //' | head -n 1)       # Mamba version (ver. 1.4.2)
yq_version=$(yq --version | sed 's/yq //')                            # Yq version (ver. 3.2.1)
rename_version="1.601"                                                # Rename version (ver. 1.601)
graphviz_version="7.1.0"                                              # GraphViz version (ver. 7.1.0)
#graphviz_version=$(dot --version | sed 's/dot - graphviz version //') # GraphViz version

fastq=$(expr $(ls -l ${workdir}/resources/reads/*.fastq.gz 2> /dev/null | wc -l)) # Get fastq.gz files count
if [[ "${fastq}" == "0" ]]                                                        # If no sample,
then                                                                               # start GeVarLi with at least 1 sample
    echo -e "${red}¡${nc} No fastq file detected in ${ylo}resources/reads/${nc} ${red}!${nc}
${red}SARS-CoV-2${nc} ${ylo}resources/data_test/${nc} fastq will be used as sample example"
    cp ${workdir}/resources/data_test/*.fastq.gz ${workdir}/resources/reads/       # using data_test/*.fastq.gz
fi
samples=$(expr ${fastq} \/ 2) # {fastq.gz count} / 2 = samples count (paired-end)

config_file="${workdir}/configuration/config.yaml"       # Get configuration file
conda_frontend=$(yq -c '.conda.frontend' ${config_file} | sed 's/\[\"//' | sed 's/\"\]//') # Get user config: conda frontend
max_threads=$(yq -r '.resources.cpus' ${config_file})    # Get user config: max threads
max_memory=$(yq -r '.resources.ram' ${config_file})      # Get user config: max memory
memory_per_job=$(expr ${max_memory} \/ ${max_threads})   # Calcul maximum memory usage per job
subset=$(yq -r '.fastq_screen.subset' ${config_file})    # Get user config: fastq_screen subsetption
time_stamp_start=$(date +"%Y-%m-%d %H:%M")               # Get system: analyzes starting time
time_stamp_archive=$(date +"%Y-%m-%d_%Hh%M")             # Convert time for archive (wo space)
SECONDS=0                                                # Initialize SECONDS counter

# Print some analyzes settings
echo -e "
${blue}Working directory${nc} _______ ${workdir}/
${blue}Samples processed${nc} _______ ${red}${samples}${nc} samples (${ylo}${fastq}${nc} fastq files)

${blue}Conda version${nc} ___________ ${ylo}${conda_version}${nc} (>= 23.3.1)
${blue}Snakemake version${nc} _______ ${ylo}${snakemake_version}${nc}
${blue}Conda frontend${nc} __________ ${ylo}${conda_frontend}${nc}
${blue}Mamba version${nc} ___________ ${ylo}${mamba_version}${nc}  

${blue}Max threads${nc} _____________ ${red}${max_threads}${nc} of ${ylo}${logical_cpu}${nc} threads available
${blue}Max memory${nc} ______________ ${red}${max_memory}${nc} of ${ylo}${ram_gb}${nc} Gb available
${blue}Jobs memory${nc} _____________ ${red}${memory_per_job}${nc} Gb per job

${blue}Fastq-Screen subset${nc} _____ ${red}${subset}${nc} reads per sample

${blue}Starting time${nc} ___________ ${time_stamp_start}
"


###############################################################################
### RENAME SAMPLES ###
######################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}RENAME FASTQ FILES${nc} ${green}#####${nc}
${green}------------------------------${nc}
"

# Rename fastq files to remove "_001" Illumina pattern (mandatory)
## De/comment line (#) if you want keep Illumina barcode-ID and/or Illumina line-ID
echo -e "Removing ${red}'_S'${nc} index tag ID"
rename "s/_S\d+_/_/" ${workdir}/resources/reads/*.fastq.gz 2> /dev/null                # Remove barcode-ID like {_S001_}
echo -e "Removing ${red}'_L'${nc} line tag ID"
rename "s/_L\d+_/_/" ${workdir}/resources/reads/*.fastq.gz 2> /dev/null                # Remove line-ID ID like {_L001_}
echo -e "Removing ${red}'_001'${nc} illumina tag ID"
rename "s/_001.fastq.gz/.fastq.gz/" ${workdir}/resources/reads/*.fastq.gz 2> /dev/null # Remove end-name ID like {_001}.fastq.gz

echo -e "
If you want to keep Illumina ${blue}barcode-ID${nc} and/or Illumina ${blue}line-ID${nc}, please edit ${ylo}Start_RQC.sh${nc} script (l.235).
"

###############################################################################
### SNAKEMAKE PIPELINES ###
###########################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}SNAKEMAKE PIPELINES${nc} ${green}#####${nc}
${green}-------------------------------${nc}
"

snakefiles_list="indexing_genomes quality_control"

echo -e "
${blue}## Unlocking Working Directory ##${nc}
${blue}---------------------------------${nc}
"
# Specify working directory (relative paths in the snakefile will use this as their origin).
# The workflow definition in form of a snakefile.
# Set or overwrite values in the workflow config object.
# Re-run all jobs the output of which is recognized as incomplete.
# Remove a lock on the working directory.

for snakefile in ${snakefiles_list} ; do
    echo -e "${blue}-- ${snakefile} --${nc}" ;
    snakemake \
	--directory ${workdir}/ \
        --snakefile ${workdir}/workflow/snakefiles/${snakefile}.smk \
        --config os=${os} \
        --rerun-incomplete \
        --unlock ;
done

echo -e "
${blue}## Conda Environments List ##${nc}
${blue}-----------------------------${nc}
"
# Specify working directory (relative paths in the snakefile will use this as their origin).
# The workflow definition in form of a snakefile.
# Use at most N CPU cores/jobs in parallel. If N is omitted or ‘all’, the limit is set to the number of available CPU cores.
# Set or overwrite values in the workflow config object.
# Re-run all jobs the output of which is recognized as incomplete.
# List all conda environments and their location on disk.

for snakefile in ${snakefiles_list} ; do
    echo -e "${blue}-- ${snakefile} --${nc}" ;
    snakemake \
        --directory ${workdir}/ \
        --snakefile ${workdir}/workflow/snakefiles/${snakefile}.smk \
        --cores ${max_threads} \
        --config os=${os} \
        --rerun-incomplete \
        --list-conda-envs ;
done

#echo -e "
#${blue}## Conda Environments Cleanup ##${nc}
#${blue}-----------------------------${nc}
#"
# Specify working directory (relative paths in the snakefile will use this as their origin).
# The workflow definition in form of a snakefile.
# Use at most N CPU cores/jobs in parallel. If N is omitted or ‘all’, the limit is set to the number of available CPU cores.
# Set or overwrite values in the workflow config object.
# Re-run all jobs the output of which is recognized as incomplete.
# Cleanup unused conda environments.
#
#for snakefile in ${snakefiles_list} ; do
#    echo -e "${blue}-- ${snakefile} --${nc}" ;
#    snakemake \
#        --directory ${workdir}/ \
#        --snakefile ${workdir}/workflow/snakefiles/${snakefile}.smk \
#        --cores ${max_threads} \
#        --config os=${os} \
#        --rerun-incomplete \
#        --conda-cleanup-envs ;
#done

echo -e "
${blue}## Conda Environments Setup ##${nc}
${blue}------------------------------${nc}
"
# Specify working directory (relative paths in the snakefile will use this as their origin).
# The workflow definition in form of a snakefile.
# Use at most N CPU cores/jobs in parallel. If N is omitted or ‘all’, the limit is set to the number of available CPU cores.
# Set or overwrite values in the workflow config object.
# Re-run all jobs the output of which is recognized as incomplete.
# If defined in the rule, run job in a conda environment.
# If mamba package manager is not available, or if you still prefer to use conda, you can enforce that with this setting.
## Default "mamba", recommended because much faster !
# If specified, only creates the job-specific conda environments then exits. The –use-conda flag must also be set.

for snakefile in ${snakefiles_list} ; do
    echo -e "${blue}-- ${snakefile} --${nc}" ;
    snakemake \
        --directory ${workdir}/ \
        --snakefile ${workdir}/workflow/snakefiles/${snakefile}.smk \
        --cores ${max_threads} \
        --config os=${os} \
        --rerun-incomplete \
        --use-conda \
        --conda-frontend ${conda_frontend} \
        --conda-create-envs-only ;
done

echo -e "
${blue}## Dry Run ##${nc}
${blue}-------------${nc}
"
# Specify working directory (relative paths in the snakefile will use this as their origin).
# The workflow definition in form of a snakefile.
# Use at most N CPU cores/jobs in parallel. If N is omitted or ‘all’, the limit is set to the number of available CPU cores.
# Set or overwrite values in the workflow config object.
# Re-run all jobs the output of which is recognized as incomplete.
# If defined in the rule, run job in a conda environment.
# If mamba package manager is not available, or if you still prefer to use conda, you can enforce that with this setting.
## Default "mamba", recommended because much faster !
# Tell the scheduler to assign creation of given targets (and all their dependencies) highest priority.
# Do not execute anything, and display what would be done. If very large workflow, use –dry-run –quiet to just print a summary of the DAG of jobs.
# Do not output any progress or rule information.

for snakefile in ${snakefiles_list} ; do
    echo -e "${blue}-- ${snakefile} --${nc}" ;
    snakemake \
        --directory ${workdir}/ \
        --snakefile ${workdir}/workflow/snakefiles/${snakefile}.smk \
        --cores ${max_threads}\
        --config os=${os} \
        --rerun-incomplete \
        --use-conda \
        --conda-frontend ${conda_frontend} \
        --dry-run \
        --quiet ;
done

echo -e "
${blue}## Let's Run! ##${nc}
${blue}----------------${nc}
"
# Specify working directory (relative paths in the snakefile will use this as their origin).
# The workflow definition in form of a snakefile.
# Use at most N CPU cores/jobs in parallel. If N is omitted or ‘all’, the limit is set to the number of available CPU cores.
# Define a global maximum number of threads available to any rule. Snakefiles requesting more threads will have their values reduced to the maximum. 
# Set or overwrite values in the workflow config object.
# Re-run all jobs the output of which is recognized as incomplete.
# Go on with independent jobs if a job fails.
# If defined in the rule, run job in a conda environment.
# If mamba package manager is not available, or if you still prefer to use conda, you can enforce that with this setting.
## Default "mamba", recommended because much faster !
# Tell the scheduler to assign creation of given targets (and all their dependencies) highest priority.
# Print out the shell commands that will be executed.

for snakefile in ${snakefiles_list} ; do
    echo -e "${blue}-- ${snakefile} --${nc}" ;
    snakemake \
        --directory ${workdir}/ \
        --snakefile ${workdir}/workflow/snakefiles/${snakefile}.smk \
        --cores ${max_threads} \
        --max-threads ${max_threads} \
        --config os=${os} \
        --rerun-incomplete \
        --keep-going \
        --use-conda \
        --conda-frontend ${conda_frontend} \
        --printshellcmds ;
done

###############################################################################
### CONCATENATE RESULTS ###
###########################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}CONCATENATE RESULTS${nc} ${green}#####${nc}
${green}-------------------------------${nc}
"

cp ${workdir}/results/00_Quality_Control/multiqc/multiqc_report.html \
   2> /dev/null \
   ${workdir}/results/All_readsQC_reports.html

		
###############################################################################
### GRAPHS, SUMMARY and LOGS ###
#############################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}GRAPHS, SUMMARY and LOGS${nc} ${green}#####${nc}
${green}------------------------------------${nc}
"

# Make directories
mkdir -p ${workdir}/results/10_Reports/graphs/ 2> /dev/null
mkdir -p ${workdir}/results/10_Reports/files-summaries/ 2> /dev/null

graph_list="dag rulegraph filegraph"
extention_list="pdf png"

for snakefile in ${snakefiles_list} ; do
    for graph in ${graph_list} ; do
	for extention in ${extention_list} ; do
	    snakemake \
		--directory ${workdir}/ \
                --snakefile ${workdir}/workflow/snakefiles/${snakefile}.smk \
                --${graph} \
	    | dot -T${extention} \
            2> /dev/null \
	    1> ${workdir}/results/10_Reports/graphs/${snakefile}_${graph}.${extention} ;
	done ;
    done ;
done

for snakefile in ${snakefiles_list} ; do
    snakemake \
        --directory ${workdir} \
        --snakefile ${workdir}/workflow/snakefiles/${snakefile}.smk \
        --summary > ${workdir}/results/10_Reports/files-summaries/${snakefile}_files-summary.txt \
    2> /dev/null ;
done

cp ${config_file} ${workdir}/results/10_Reports/config.log 2> /dev/null

###############################################################################
### CLEAN and SAVE ###
######################
echo -e "
${green}------------------------------------------------------------------------${nc}
${green}#####${nc} ${red}CLEAN & SAVE${nc} ${green}#####${nc}
${green}------------------------${nc}
"

# Save and deactive environments
mkdir -p ${workdir}/results/10_Reports/conda_env/ 2> /dev/null
cp ${workdir}/workflow/environments/${os}/*.yaml ${workdir}/results/10_Reports/conda_env/
conda deactivate

# Cleanup
find ${workdir}/results/ -type f -empty -delete # Remove empty file (like empty log)
find ${workdir}/results/ -type d -empty -delete # Remove empty directory

# Timer
time_stamp_end=$(date +"%Y-%m-%d %H:%M") # Get date / hour ending analyzes
elapsed_time=${SECONDS}                  # Get SECONDS counter 
minutes=$((${elapsed_time}/60))          # / 60 = minutes
seconds=$((${elapsed_time}%60))          # % 60 = seconds

# Print timer
echo -e "
${blue}Start time${nc} _____________ ${time_stamp_start}
${blue}End time${nc} _______________ ${time_stamp_end}
${blue}Processing time${nc} ________ ${ylo}${minutes}${nc} minutes and ${ylo}${seconds}${nc} seconds elapsed
"

# Log analyzes settings
echo "
Name ___________________ Start_RQC.sh
Version ________________ ${rqc_version}
Author _________________ Nicolas Fernandez
Affiliation ____________ IRD_U233_TransVIHMI
Aim ____________________ Bash script for RQC
Date ___________________ 2021.10.12
Latest modifications ___ 2023.06.03
Run ____________________ bash Start_RQC.sh

Operating System _______ ${os}
Shell __________________ ${shell}

                   Brand(R) | Type(R) | Model | @ Speed GHz
Chip Model Name ________ ${model_name}
Physical CPUs __________ ${physical_cpu} cores
Logical CPUs ___________ ${logical_cpu} threads
System Memory __________ ${ram_size} Gb of RAM

Working directory ______ ${workdir}/
Samples processed ______ ${samples} samples (${fastq} fastq files)

Conda version __________ ${conda_version}
Snakemake version ______ ${snakemake_version}
Conda frontend _________ ${conda_frontend}
Mamba version __________ ${mamba_version}

Max threads ____________ ${max_threads} of ${logical_cpu} threads available
Max memory _____________ ${max_memory} of ${ram_size} Gb available
Jobs memory ____________ ${memory_per_job} Gb per job maximum
Fastq-Screen subset _____ ${subset} reads per sample

Start time _____________ ${time_stamp_start}
End time _______________ ${time_stamp_end}
Processing time ________ ${minutes} minutes and ${seconds} seconds elapsed
" > ${workdir}/results/10_Reports/settings.log

# Gzip reports directory
cd ${workdir}/results/
tar -zcf 10_Reports_archive.tar.gz 10_Reports

# Gzip results directory
mkdir -p ${workdir}/archives/ 2> /dev/null
cd ${workdir}/
tar -zcf archives/Results_${time_stamp_archive}_${samples}sp_archive.tar.gz results/

echo -e "
${green}------------------------------------------------------------------------${nc}
"
###############################################################################