Commit 93cf42e8 authored by nchernia's avatar nchernia
Browse files

Merge branch 'encode' of https://github.com/aidenlab/juicer into encode

parents a50c79f1 6cb20e29

CPU/common/juicer_hiccups.sh

+2 −2
Changes for CPU/common/juicer_hiccups.sh: 2 added lines, 2 removed lines.
Original line number Diff line number Diff line
@@ -78,14 +78,14 @@ fi 
if [ -e ${hic_file_path%.*}"_loops" ]

then

    echo -e "\nAPA:\n"

    ${juicer_tools_path} apa ${hic_file_path} ${hic_file_path%.*}"_loops" "apa_results"

    ${juicer_tools_path} apa ${hic_file_path} ${hic_file_path%.*}"_loops/merged_loops.bedpe" "apa_results"

    ## Check that bed folder exists    

    if [ ! -e "${bed_file_dir}" ]; then

	echo "***! Can't find folder ${bed_file_dir}";

	echo "***! Not running motif finder";

    else

	echo -e "\nMOTIF FINDER:\n"

	${juicer_tools_path} motifs ${genomeID} ${bed_file_dir} ${hic_file_path%.*}"_loops"

	${juicer_tools_path} motifs ${genomeID} ${bed_file_dir} ${hic_file_path%.*}"_loops/merged_loops.bedpe"

    fi

    echo -e "\n(-: Feature annotation successfully completed (-:"

else

CPU/juicer.sh

+1 −1
Changes for CPU/juicer.sh: 1 added line, 1 removed line.
Original line number Diff line number Diff line
@@ -593,7 +593,7 @@ if [ -z $postproc ] 
    export IBM_JAVA_OPTIONS="-Xmx1024m -Xgcthreads1"

    export _JAVA_OPTIONS="-Xmx1024m -Xms1024m"

    tail -n1 $headfile | awk '{printf"%-1000s\n", $0}' > $outputdir/inter.txt

    dups=$(samtools view -c -f 1089 -F 256 $outputdir/merged_dedup.bam)

    dups=$(samtools view -c -f 1089 -F 256 $sthreadstring $outputdir/merged_dedup.bam)

    cat $splitdir/*.res.txt | awk -v ligation=$ligation -v dups=$dups -f ${juiceDir}/scripts/common/stats_sub.awk >> $outputdir/inter.txt

    cp $outputdir/inter.txt $outputdir/inter_30.txt

SLURM/scripts/juicer.sh

+29 −5
Changes for SLURM/scripts/juicer.sh: 29 added lines, 5 removed lines.
Original line number Diff line number Diff line
@@ -104,7 +104,7 @@ then 
    # Juicer directory, contains scripts/, references/, and restriction_sites/

    # can also be set in options via -D

    # XXX verify / change me

    juiceDir="/gpfs0/juicer"

    juiceDir="/gpfs0/juicer2"

    load_bwa="unset PYTHONPATH; conda activate"

    # default queue, can also be set in options via -q

    queue="mhgcp"
@@ -122,7 +122,7 @@ else 
    load_samtools="spack load samtools arch=\`spack arch\`"

    # Juicer directory, contains scripts/, references/, and restriction_sites/

    # can also be set in options via -D

    juiceDir="/gpfs0/juicer"

    juiceDir="/gpfs0/juicer2"

    # default queue, can also be set in options

    queue="commons"

    queue_time="2880"
@@ -163,9 +163,11 @@ justexact=0 
assembly=0

# force cleanup

cleanup=0

# qc apa

qc_apa=0



## Read arguments                                                     

usageHelp="Usage: ${0##*/} [-g genomeID] [-d topDir] [-q queue] [-l long queue] [-s site]\n                 [-a about] [-S stage] [-p chrom.sizes path]\n                 [-y restriction site file] [-z reference genome file]\n                 [-C chunk size] [-D Juicer scripts directory]\n                 [-Q queue time limit] [-L long queue time limit] [-b ligation] [-t threads] [-T threadsHic]\n                 [-A account name] [-e] [-h] [-f] [-j] [--assembly] [--cleanup]"

usageHelp="Usage: ${0##*/} [-g genomeID] [-d topDir] [-q queue] [-l long queue] [-s site]\n                 [-a about] [-S stage] [-p chrom.sizes path]\n                 [-y restriction site file] [-z reference genome file]\n                 [-C chunk size] [-D Juicer scripts directory]\n                 [-Q queue time limit] [-L long queue time limit] [-b ligation] [-t threads] [-T threadsHic]\n                 [-A account name] [-e] [-h] [-f] [-j] [--assembly] [--cleanup] [--qc_apa]"

genomeHelp="* [genomeID] must be defined in the script, e.g. \"hg19\" or \"mm10\" (default \n  \"$genomeID\"); alternatively, it can be defined using the -z command"

dirHelp="* [topDir] is the top level directory (default\n  \"$topDir\")\n     [topDir]/fastq must contain the fastq files\n     [topDir]/splits will be created to contain the temporary split files\n     [topDir]/aligned will be created for the final alignment"

queueHelp="* [queue] is the queue for running alignments (default \"$queue\")"
@@ -244,6 +246,7 @@ while getopts "d:g:a:hq:s:p:l:y:z:S:C:D:Q:L:b:A:i:t:jfec-:T:" opt; do 
	-) case "${OPTARG}" in 

	    assembly) earlyexit=1; assembly=1 ;;

	    cleanup)  cleanup=1 ;;

	    qc_apa)   qc_apa=1 ;;

	    *) printHelpAndExit 1;;

           esac;;

	[?]) printHelpAndExit 1;;
@@ -1277,7 +1280,7 @@ FINCLN1` 
	    time ${juiceDir}/scripts/juicer_tools pre -n -f $site_file -s $outputdir/inter.txt -g $outputdir/inter_hists.m -q 1 -r 2500000,1000000,500000,250000,100000,50000,25000,10000,5000,2000,1000,500,200,100 $threadHicString $outputdir/merged1.txt $outputdir/inter.hic $genomePath

	fi

	time ${juiceDir}/scripts/juicer_tools addNorm $threadNormString ${outputdir}/inter.hic 

	rm -R ${outputdir}"/HIC_tmp"

	rm -Rf ${outputdir}"/HIC_tmp"

	date

HIC`
@@ -1319,7 +1322,7 @@ HIC` 
	    time ${juiceDir}/scripts/juicer_tools pre -n -f $site_file -s $outputdir/inter_30.txt -g $outputdir/inter_30_hists.m -q 30 -r 2500000,1000000,500000,250000,100000,50000,25000,10000,5000,2000,1000,500,200,100 $threadHic30String $outputdir/merged30.txt $outputdir/inter_30.hic $genomePath

	fi

	time ${juiceDir}/scripts/juicer_tools addNorm $threadNormString ${outputdir}/inter_30.hic

	rm -R ${outputdir}"/HIC30_tmp"

	rm -Rf ${outputdir}"/HIC30_tmp"

	date

HIC30`
@@ -1389,6 +1392,27 @@ jid=`sbatch <<- ARROWS | egrep -o -e "\b[0-9]+$" 
	date;

ARROWS`

dependarrows="${dependhiccups}:$jid"

if [ "$qc_apa" = 1 ]

then

    jid=`sbatch <<- QC | egrep -o -e "\b[0-9]+$"

	#!/bin/bash -l

	#SBATCH -p $queue

	#SBATCH --mem-per-cpu=4G

	#SBATCH -o $debugdir/qc-%j.out

	#SBATCH -e $debugdir/qc-%j.err

	#SBATCH -t $queue_time

	#SBATCH --ntasks=1

	#SBATCH -J "${groupname}_qc_apa"

	${sbatch_wait}

	$userstring

	${load_java}

	date

	export IBM_JAVA_OPTIONS="-Xmx4000m -Xgcthreads1"                                                                    

        export _JAVA_OPTIONS="-Xmx4000m -Xms4000m" 

	java -jar ${juiceDir}/scripts/juicer_3.25.21_aggNormAPA.jar apa --ag-norm -k NONE -n 300 -w 100 -r 1000 -q 60 --threads 1 $outputdir/inter_30.hic ${juiceDir}/scripts/GSE63525_GM12878_primary+replicate_HiCCUPS_looplist_with_motifs_unique_localized.txt $outputdir/qc_apa

	date

QC`

fi



jid=`sbatch <<- FINCLN1 | egrep -o -e "\b[0-9]+$"

	#!/bin/bash -l

SLURM/scripts/juicer_hiccups.sh

+2 −2
Changes for SLURM/scripts/juicer_hiccups.sh: 2 added lines, 2 removed lines.
Original line number Diff line number Diff line
@@ -95,14 +95,14 @@ fi 
if [ -n "$(ls -A ${hic_file_path%.*}_loops)" ]

then

    echo -e "\nAPA:\n"

    ${juicer_tools_path} apa ${hic_file_path} ${hic_file_path%.*}"_loops" "apa_results"

    ${juicer_tools_path} apa ${hic_file_path} ${hic_file_path%.*}"_loops/merged_loops.bedpe" "apa_results"

    ## Check that bed folder exists    

    if [ ! -e "${bed_file_dir}" ]; then

	echo "***! Can't find folder ${bed_file_dir}";

	echo "***! Not running motif finder";

    else

	echo -e "\nMOTIF FINDER:\n"

	${juicer_tools_path} motifs ${genomeID} ${bed_file_dir} ${hic_file_path%.*}"_loops"

	${juicer_tools_path} motifs ${genomeID} ${bed_file_dir} ${hic_file_path%.*}"_loops/merged_loops.bedpe"

    fi

    echo -e "\n(-: Feature annotation successfully completed (-:"

else

SLURM/scripts/mega.sh

+3 −4
Changes for SLURM/scripts/mega.sh: 3 added lines, 4 removed lines.
Original line number Diff line number Diff line
@@ -73,7 +73,7 @@ else 
    load_gpu="CUDA_VISIBLE_DEVICES=0,1,2,3"

    # Juicer directory, contains scripts/, references/, and restriction_sites/

    # can also be set in options via -D

    juiceDir="/gpfs0/juicer/"

    juiceDir="/gpfs0/juicer2/"

    # default queue, can also be set in options

    queue="commons"

    queue_time="1400"
@@ -187,7 +187,6 @@ touchfile3=${megadir}/touch3 
touchfile4=${megadir}/touch4



## Check for existing merged files:



merged_count=`find -L ${topDir} | grep merged1.txt | wc -l`

if [ "$merged_count" -lt "1" ]

then
@@ -442,7 +441,7 @@ fi 
	    time ${juiceDir}/scripts/juicer_tools pre -n -f $site_file -s $outputdir/inter.txt -g $outputdir/inter_hists.m -q 1 -r 2500000,1000000,500000,250000,100000,50000,25000,10000,5000,2000,1000,500,200,100 $threadHicString $outputdir/merged1.txt $outputdir/inter.hic $genomeID

	fi

	time ${juiceDir}/scripts/juicer_tools addNorm $threadNormString ${outputdir}/inter.hic 

	rm -R ${outputdir}"/HIC_tmp"

	rm -Rf ${outputdir}"/HIC_tmp"

HIC1`

    dependhic1="afterok:$jid5"
@@ -481,7 +480,7 @@ fi 
	    time ${juiceDir}/scripts/juicer_tools pre -n -f $site_file -s $outputdir/inter_30.txt -g $outputdir/inter_30_hists.m -q 30 -r 2500000,1000000,500000,250000,100000,50000,25000,10000,5000,2000,1000,500,200,100 $threadHic30String $outputdir/merged30.txt $outputdir/inter_30.hic $genomeID

	fi

	time ${juiceDir}/scripts/juicer_tools addNorm $threadNormString ${outputdir}/inter_30.hic

	rm -R ${outputdir}"/HIC30_tmp"

	rm -Rf ${outputdir}"/HIC30_tmp"

HIC30`

    dependhic30only="afterok:$jid6"

    sbatchdepend="#SBATCH -d ${dependhic30only}"

CRA Git | Maintained and supported by SUSTech CRA and CCSE