New upstream version 2.4.3

################################################################################

SET (gatb-tool_VERSION_MAJOR 2)

SET (gatb-tool_VERSION_MINOR 4)

SET (gatb-tool_VERSION_PATCH 2)

SET (gatb-tool_VERSION_PATCH 3)

IF (DEFINED MAJOR)

    SET (gatb-tool_VERSION_MAJOR ${MAJOR})

)

# For creating the BINARY package we include the files we want

INSTALL (DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/discoSnpRAD        DESTINATION .)

INSTALL (DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/doc                DESTINATION .)

INSTALL (DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/test               DESTINATION .)

INSTALL (DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/scripts            DESTINATION . FILES_MATCHING REGEX ".*\\.(py|sh)$"  PATTERN "jenkins" EXCLUDE)

## Installation

* discoSnp++

* discoSnp++ (see [../README.md](../README.md))

* `short_read_connector` must have been downloaded and installed (clustering task). [https://github.com/GATB/short_read_connector](https://github.com/GATB/short_read_connector)

## Usage

```

./run_discoSnpRad.sh --fof read_file_of_files --src_path <directory> [discoSnp++ OPTIONS]

./run_discoSnpRad.sh -r read_file_of_files -S -p myDiscoSnpRADresult [discoSnp++ OPTIONS]

```

Clustering

**Clustering option** (RAD-specific option):

```

-S|--src_path <directory>

    **absolute** path to short_read_connector directory, containing the "short_read_connector.sh" file. 

-S|--src [src_path]

    performs clustering of variants with short_read_connector

    src_path: **absolute** path to short_read_connector directory, containing the "short_read_connector.sh" file. 

    -Note1: short read connector must be compiled.

    -Note2: with this option, discoSnpRad provide a vcf file containing SNPs and INDELS, clustered by locus

    -Note2: if no value is given, it assumes short_read_connector.sh is in the PATH env variable.

    -Note3: with this option, discoSnpRad outputs a vcf file containing the variants clustered by locus.

```

All other options are described in [discoSnp++ README](../README.md). Note that many discoSNP++ parameters have here default values, specifically adapted to RAD-seq data.

## Output

* a log file reminds all filtering steps applied and the name of the output .vcf file

* a vcf file containing results of filtering and clustering

When run with output prefix name `myDiscoSnpRADresult`, the main output file is :

* `myDiscoSnpRADresult_[parameter_values]_clustered.vcf`: the final set of variants, with various information, including clustering per locus information (see VCF format below).

* or `myDiscoSnpRADresult_[parameter_values].vcf` if no clustering was performed.

Additionnally, several other files are output that can be usefull :

* `myDiscoSnpRADresult_[parameter_values]_raw.fa`: the raw set of variants in fasta format, prior to any filtering and clustering steps.

* `myDiscoSnpRADresult_[graph_parameter_values].h5`: the de Bruijn graph in h5 format (reusable with any GATB tool)

* `myDiscoSnpRADresult_read_files_correspondance.txt`: the correspondence between read file names and IDs given as genotypes in the vcf

* the standard output reminds all filtering steps applied and the name of the output .vcf file

#### VCF format

Each variant is described with: 

* an ID: `ID` column, 

* two alleles (`REF` and `ALT` columns), 

* a quality value: `INFO` column, `Rk`, between 0 (bad) and 1 (best),

* some clustering information: `INFO` field: with the locus id (`Cluster`) and its number of varying sites (`ClSize`),

* and for each sample in the genotype columns (`G1`, `G2`,...): the inferred genotype (`0/0`, `0/1`, `1/1`or `./.`for missing value), the read depths (`RD` total, `AD`per allele), among others.

## Content of this directory

echo " 1/ Remove variants with more than 95% missing genotypes and low rank (<0.4)"

echo " 2/ Cluster variants per locus"

echo " 3/ Format the variants in a vcf file with cluster information"

echo "Usage: ./discoRAD_clustering.sh -f discofile -s SRC_directory/ -o output_file.vcf"

echo "Usage: ./discoRAD_clustering.sh -f discofile -s SRC_path -o output_file.vcf"

# echo "nb: all options are MANDATORY\n"

echo "OPTIONS:"

echo "\t -f: DiscoSnp fasta output containing coherent predictions"

        ;;

        s)

        short_read_connector_directory=$OPTARG

        short_read_connector_path=$OPTARG

        ;;

        o)

    echo "${red}-f is mandatory$reset" >&2

    exit

fi

if [[ -z "${short_read_connector_directory}" ]]; then

if [[ -z "${short_read_connector_path}" ]]; then

    echo "${red}-s is mandatory$reset" >&2

    exit

fi

if [ -d "$EDIR/../../build/" ] ; then # VERSION SOURCE COMPILED

BINDIR=$EDIR"/../../build/bin"

else # VERSION BINARY

BINDIR=$EDIR"/../../bin"

BINDIR=$EDIR/../../bin

fi

rawdiscofile_base=$( basename "${rawdiscofile}" .fa)

#ls ${disco_simpler}.fa > ${disco_simpler}.fof

# Compute sequence similarities

cmdSRC="${short_read_connector_directory}/short_read_connector.sh -b ${disco_simpler}.fa -q ${disco_simpler}.fof -s 0 -k ${usedk} -a 1 -l -p ${disco_simpler}  1>&2 "

cmdSRC="${short_read_connector_path} -b ${disco_simpler}.fa -q ${disco_simpler}.fof -s 0 -k ${usedk} -a 1 -l -p ${disco_simpler}  1>&2 "

echo $green$cmdSRC$cyan

if [[ "$wraith" == "false" ]]; then

    eval $cmdSRC

genotyping="-genotype"

remove=1

verbose=1

clustering="false"

short_read_connector_path=""

option_phase_variants=""

#EDIR=$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )

    echo " ************"

    echo "run_discoSnpRad.sh, pipelining kissnp2 and kissreads and clustering per locus for calling SNPs and small indels from RAD-seq data without the need of a reference genome"

    echo "Version "$version

    echo "Usage: ./run_discoSnpRad.sh --fof read_file_of_files --src_path <directory> [OPTIONS]"

    echo "Usage: ./run_discoSnpRad.sh --fof read_file_of_files --src [src_path] [OPTIONS]"

    echo -e "MANDATORY"

    echo -e "\t -r|--fof <file name of a file of file(s)>"

    echo -e "\t\t The input read files indicated in a file of file(s)"

    echo -e "\t\t Example: -r bank.fof with bank.fof containing the two lines \n\t\t\t data_sample/reads_sequence1.fasta\n\t\t\t data_sample/reads_sequence2.fasta.gz"

    echo -e "\nOPTIONS"

    echo -e "\t -S|--src_path <directory>"

    echo -e "\t\t **absolute** path to short_read_connector directory, containing the \"short_read_connector.sh\" file. "

    echo -e "\t -S|--src [src_path]"

    echo -e "\t\t performs clustering of variants with short_read_connector"

    echo -e "\t\t src_path: **absolute** path to short_read_connector directory, containing the \"short_read_connector.sh\" file. "

    echo -e "\t\t -Note1: short read connector must be compiled."

    echo -e "\t\t -Note2: with this option, discoSnpRad provide a vcf file containing SNPs and INDELS, clustered by locus" 

    echo -e "\t\t -Note2: if no value is given, it assumes short_read_connector.sh is in the PATH env variable."

    echo -e "\t\t -Note3: with this option, discoSnpRad outputs a vcf file containing the variants clustered by locus" 

    echo -e "\t -k | --k_size value <int value>"

    echo -e "\t\t Set the length of used kmers. Must fit the compiled value."

    -w)

        wraith="true"

        ;;

    -S|--src_path)

    -S|--src)

    	clustering="true"

        if [ "$2" ] && [ ${2:0:1} != "-" ] ; then # checks that there exists a second value and its is not the start of the next option

            short_read_connector_path=$2

            shift

        else

            die 'ERROR: "'$1'" option requires a non-empty option argument.'

        fi

        ;;

    -a|--ambiguity_max_size)

        if [ "$2" ] && [ ${2:0:1} != "-" ] ; then # checks that there exists a second value and its is not the start of the next option

            max_ambigous_indel=$2

    exit 1

fi

src_file="$short_read_connector_path/short_read_connector.sh"

if [[ "$wraith" == "false" ]]; then

    echo $yellow${src_file}$reset

fi

#Checks if clustering can be performed

if [[ "$wraith" == "false" ]]; then

if [[ "$clustering" == "true" ]]; then

	# first tests the directory given by user if any

	if [ -n "$short_read_connector_path" ]; then

		src_file="$short_read_connector_path/short_read_connector.sh"

    	if [ -f "$src_file" ]; then

        if [[ "$wraith" == "false" ]]; then

            echo "${yellow}short_read_connector is $src_file$reset"

        fi

            echo "${yellow}short_read_connector path is $src_file$reset"

    	else

        if [[ "$wraith" == "false" ]]; then

    		echo -e "${red}\t\t\t**************************************************************************"

            echo -e "\t\t\t** WARNING: I cannot find short_read_connector (-S). "

            echo -e "\t\t\t** $src_file does not exist"

            echo -e "\t\t\t** I will not cluster variants per RAD locus"

            echo -e "\t\t\t**************************************************************************"

            echo $reset

    		clustering="false"

    	fi

    else

    	#then tests if src is in the PATH env variable

    	src_file=$(command -v short_read_connector.sh)

    	if [ -n "$src_file" ]; then

    		echo "${yellow}short_read_connector path is $src_file$reset"

    	else

    		echo -e "${red}\t\t\t**************************************************************************"

            echo -e "\t\t\t** WARNING: I cannot find short_read_connector in PATH. "

            echo -e "\t\t\t** Try giving the absolute path of short_read_connector directory with option -S"

            echo -e "\t\t\t** I will not cluster variants per RAD locus"

            echo -e "\t\t\t**************************************************************************"

            echo $reset

    		clustering="false"

    	fi

    fi

fi

######### CHECK THE k PARITY ##########

rest=$(( $k % 2 ))

if [ $rest -eq 0 ]

echo -e "\t###############################################################$reset"

T="$(date +%s)"

if [ -f "$src_file" ]; then

if [[ "$clustering" == "true" ]]; then

    if [[ "$wraith" == "false" ]]; then

        echo "${yellow}Clustering and vcf formmatting$reset"

    fi

    final_output="${kissprefix}_clustered.vcf"

    cmd="$EDIR/clustering_scripts/discoRAD_clustering.sh -f ${kissprefix}_raw.fa -s $short_read_connector_path -o ${final_output}"

    echo $green$cmd$cyan

    cmd="$EDIR/clustering_scripts/discoRAD_clustering.sh -f ${kissprefix}_raw.fa -s $src_file -o ${final_output}"

    echo $green$cmd$cyan$reset

    if [[ "$wraith" == "false" ]]; then

        eval $cmd

    fi  

    fi

    final_output="${kissprefix}.vcf"

    cmd="python3 $EDIR/../scripts/create_filtered_vcf.py -i ${kissprefix}_raw.fa -o ${final_output} -m 0.95 -r 0.4"

    echo $green$cmd$cyan

    echo $green$cmd$cyan$reset

    if [[ "$wraith" == "false" ]]; then

        eval $cmd

    fi