Directions:
Go to the run directory.
git clone https://github.com/vari-bbc/demultiplex.git
OPTIONAL: Add the directive for email notifications in demultiplex/bcl2fastq_snake.sh file (#PBS -M [email protected])
sbatch -p genomics demultiplex/bcl2fastq_snake.sh
Directions:
Go to the run directory.
git clone https://github.com/vari-bbc/demultiplex.git
OPTIONAL: change the email in demultiplex/bcl2fastq.sh file (#PBS -M [email protected])
sbatch -p genomics demultiplex/bcl2fastq.sh
If this doesn't work, check the demultiplex_workflow.[oe]JOB.ID files that are in the run directory.
If merging lanes failed (e.g. some samples did not get demultiplexed properly), but you still want to proceed with the rest of the pipeline, then in the run directory type 'touch mergelanes.override' <- this will create a file 'mergelanes.override' that forces the pipeline to continue with the remaining samples. You can specify this before the run if you want.
Enable or disable running Preseq, by changing 'run' to True or False in bcl2fastq_snake/config.yaml. You must also indicate whether the data is 'DNA' or 'RNA' sequencing and which reference to align to. Look inside bcl2fastq_snake/config.yaml for more details.
Empty placeholder fastq files and output files from other QC tools will be created for samples with no reads. Fastq files that end up with 0 reads will be listed in a file named missing_fastqs.log immediately after bcl2fastq is run and will also be listed in the multiQC reports.
Low count files, which are fastq files with read count less than the minimum value set in demultiplex/bcl2fastq_snake/config.yaml, are listed in low_count_fastqs.log only after all FastQC jobs are finished because it uses the FastQC output to determine the counts. These low count files are also listed in the multiQC reports.
If a samplesheet error is noticed after examining missing_fastqs.log or low_count_fastqs.log, you may want to kill all the demultiplex PBS jobs immediately instead of waiting for the jobs to finish. After that, you can 'reset' the demultiplexing directory and restart the pipeline by running qsub -q genomics demultiplex/bcl2fastq_snake.sh (if using the Snakemake pipeline) again. See sections below on killing PBS jobs and how to 'reset' the directory.
If you want to stop the demultiplexing pipeline, you need to kill all the running jobs.
First, run the following to check that only demultiplexing jobs are matched.
qstat -u user.name | grep -P '\ssnakejob\.|demultiplex'
Then, run the following to kill those jobs.
qstat -u user.name | grep -P '\ssnakejob\.|demultiplex' | grep -Po '^\d+' | xargs qdel
In case you need to rerun a pipeline, you may want to get rid of files created by the previous run.
./demultiplex/reset_dmux.sh
This can take some time to run and uses up to 8 cores, so you should run this as an interactive PBS job, requesting 8 cores.
./demultiplex/fastq_md5sums.sh
Directions:
Go to the run directory.
git clone https://github.com/ianbed/demultiplex.git
OPTIONAL: change the email in demultiplex/cellranger_atac_demultiplex.sh file (#PBS -M [email protected])
qsub -q genomics demultiplex/cellranger_atac_demultiplex.sh
If this doesn't work, check the sc-atac-demux.[oe]JOB.ID files that are in the run directory.