snake_hisat

Workflow for RNAseq, using hisat2 aligner

get the hisat index for human

wget https://cloud.biohpc.swmed.edu/index.php/s/grch38/download
mv download grch39.tar.gz
tar -xvzf grch39.tar.gz

link the location in the Snakefile

eg

GENOME="/cluster/home/michalo/project_michalo/hisat/grch38/genome"

get the GTF

wget ftp://ftp.ensembl.org/pub/release-99/gtf/homo_sapiens/Homo_sapiens.GRCh38.99.gtf.gz
gunzip Homo_sapiens.GRCh38.99.gtf.gz

link the GTF in Snakefile

eg

GTF="/cluster/home/michalo/project_michalo/hg38/Homo_sapiens.GRCh38.99.gtf"

Software required:

If you want to use it locally, the software from the workflow: trimmomatic, hisat, subread, samtools need to be installed locally and made runnable from command line

Adapting

The paths to genome, GTF and adapters need to be set in the python constants in the Snakefile If needed, also paths to the software commands and trimmomatic jar. Recommended is to have them in the executable or java paths, eg with setting the environment value.

Running

Create a run directory, where you place: Snakefile, adapters.fa and fastq.gz files in "data" subdirectory. Do the updates to the Snakefile as above: location of genome index and GTF annotation, then:

dry run

snakemake -np

normal run

snakemake -p

run on the cluster

Make the snakemake available in the cluster environment, eg

module load gcc/8.2.0 python/3.10.4

LSF

snakemake -p -j 999 --cluster-config cluster.json --cluster "bsub -W {cluster.time} -n {cluster.n}"

SLURM

# change times in cluster.json to HH:MM:SS
snakemake -p -j 999 --cluster-config cluster.json --cluster "sbatch --time {cluster.time} -n {cluster.n}"
snakemake -p -j 999 --cluster-config cluster.json --cluster "sbatch --time {cluster.time} -n 1 --cpus-per-task={cluster.n}"
snakemake -p -j 999 --cluster-config cluster.json --cluster "sbatch --time {cluster.time} -n 1 --cpus-per-task={cluster.n} --mem-per-cpu={cluster.mem}"

SLURM with containers

Running the workflow with the containers from Galaxy software stack requires passing the external folders as singularity parameters to the snakemake. The containers will be loaded into .snakemake folder.

 snakemake -p -j 999 --use-singularity --cluster-config cluster.json \
  --cluster "sbatch --time {cluster.time} -n 1 --cpus-per-task={cluster.n}" \
  --singularity-args "--bind /cluster/scratch/username/runfolder/:/mnt2 --bind /cluster/home/michalo/project_michalo/hisat/grch38/:/genomes --bind /cluster/home/michalo/project_michalo/hg38/:/annots"