set up for the Juicer genome draft assembly on HPC

[Nguyen Xuan-Thang]

Hi Juicer team,

Thank you for the maintain and active resolve the bugs

I am new with Juicer and/or HiC analysis. I wonder where is my mistakes when I set up the run on HPC
I install all dependency in conda env Juicer_env and make changes the folder and parameters in juicer.sh, however, I have no idea why it still got stucks somewhere.
 

Errors:
${juicer_outdir}/scripts/juicer.sh         -g draft         -t 72         -s HindIII         -z ${juicer_outdir}/references/LIPO863B_NextDenovo_NextPolish.fasta         -y ${juicer_outdir}/restriction_sites/LIPO863B_NextDenovo_draft_HindIII.txt         -p assembly         -D ${juicer_outdir}/
Using /staging/leuven/stg_00096/home/thangnx/LPS_WGS_shortread/HiC/opt/juicer/restriction_sites/LIPO863B_NextDenovo_draft_HindIII.txt as site file
(-: Looking for fastq files...fastq files exist
(-: Aligning files matching /staging/leuven/stg_00096/home/thangnx/LPS_WGS_shortread/HiC/opt/juicer/fastq/*_R*.fastq*
 in queue batch to genome draft with no fragment delimited maps.
---  Using already created files in /staging/leuven/stg_00096/home/thangnx/LPS_WGS_shortread/HiC/opt/juicer/splits

(-: Starting job to launch other jobs once splitting is complete
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
sbatch: error: Batch job submission failed: Job dependency problem
(-: Finished adding all jobs... Now is a good time to get that cup of coffee... Last job id


slurm jobs:

#SBATCH --job-name=HiC_LPS_celline

# Project:

# Wall clock limit:

#SBATCH -n 1

#SBATCH -c 36

#SBATCH --mem 200G

#SBATCH --time=2-00:00:00

# Number of cores, nodes and CPUs

#SBATCH --account lp_lig

#SBATCH --cluster wice

#SBATCH --mail-type=ALL

#SBATCH --output=20250506_WGS_HiC_Assembly_LIPO863B.slurm.xxxxxxx.out

#SBATCH --mail-user=xuanthan...@student.kuleuven.be

source miniconda3/bin/activate juicer_HiC

juicer_outdir="/HiC/SLURM"

${juicer_outdir}/scripts/juicer.sh \

-g draft \

-t 72 \

-s HindIII \

-z ${juicer_outdir}/references/LIPO863B_NextDenovo_NextPolish.fasta \

-y ${juicer_outdir}/restriction_sites/LIPO863B_NextDenovo_draft_HindIII.txt \

-p assembly \

-D ${juicer_outdir}/

[Nguyen Xuan-Thang]

Here is my change for the code of juicer.sh, but the errors is jobdependency

shopt -s extglob

juicer_version="1.6"

## Set the following variables to work with your system

# Aiden Lab specific check

isRice=$(host $(hostname) | awk '{if ($1~/rice/){print 1}else {print 0}}') #'

isBCM=$(host $(hostname) | awk '{if ($1~/bcm/){print 1}else {print 0}}') #'

isVoltron=0

## path additionals, make sure paths are correct for your system

## use cluster load commands

if [ $isRice -eq 1 ]

then

export PATH=$HOME/bin:$PATH

isNots=$(host $(hostname) | awk '{if ($1~/nots/){print 1}else {print 0}}') #'

if [ $isNots -eq 1 ]

then

load_bwa="module load BWA"

load_java="module load Java/1.7.0_80"

load_gpu="module load CUDA/7.5.18"

else

load_bwa="BWA/0.7.17-foss-2018a"

load_java="module load Java/1.8.0_162"

load_gpu="module load CUDA/7.5.18"

fi

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

# can also be set in options via -D

juiceDir="/staging/leuven/stg_00096/home/thangnx/LPS_WGS_shortread/HiC/opt/juicer" ### RICE

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

queue="batch"

queue_time="12:00:00"

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

long_queue="batch"

long_queue_time="72:00:00"

elif [ $isBCM -eq 1 ]

then

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

# can also be set in options via -D

juiceDir="/staging/leuven/stg_00096/home/thangnx/LPS_WGS_shortread/HiC/opt/juicer"

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

queue="batch"

queue_time="12:00:00"

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

long_queue="batch"

long_queue_time="72:00:00"

else

isVoltron=1

# Use conda-based environment

conda_env_path="/staging/leuven/stg_00096/home/thangnx/miniconda3"

load_conda_env="source /staging/leuven/stg_00096/home/thangnx/miniconda3/bin/activate juicer_HiC"

# unset MALLOC_ARENA_MAX # on IBM platform this parameter has significant speed efect but may result in memory leaks

load_bwa="module load BWA"

load_awk="module load gawk/5.3.0-GCC-13.2.0"

load_gpu="module load CUDA/7.5.18"

call_gem="/staging/leuven/stg_00096/home/thangnx/software/gem3-mapper/bin/gem-mapper --3c"

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

# can also be set in options via -D

juiceDir="/staging/leuven/stg_00096/home/thangnx/LPS_WGS_shortread/HiC/opt/juicer"

# default queue, can also be set in options

queue="batch"

queue_time="24:00:00"

# default long queue, can also be set in options

long_queue="batch"

long_queue_time="72:00:00"

fi