Ray is a parallel software that computes de novo genome assemblies with next-generation sequencing data. Ray is written in C++ and can run in parallel on numerous interconnected computers using the message-passing interface (MPI) standard.
See versions of ray which are available:
$ module avail ray
See the built-in help:
$ Ray -help | less
Running Ray on two different nodes via MPI with a total of 15 CPUs:
#!/bin/env bash #SBATCH -p batch #SBATCH -n 15 #SBATCH -J ray #SBATCH --nodes=2 module load ray/2.3.1 readonly DATADIR=~/data/bfusca readonly OUTDIR=~/data/bfusca/ray-sept26 # mpirun gets the number of processors from SLURM mpirun Ray -k 27 -p ${DATADIR}/1_S1_L001_R1_001_val_1.fastq.gz ${DATADIR}/1_S1_L001_R2_001_val_2.fastq.gz -o ${OUTDIR}
Save in a file and submit using sbatch ray-mpi.sbatch
.
Notes from the sysadmin during installation:
$ cd /tmp $ git clone https://github.com/sebhtml/RayPlatform.git $ git clone https://github.com/sebhtml/ray.git $ cd ray $ git checkout v2.3.1 $ scl enable devtoolset-1.1 bash $ module load openmpi/1.8.2 $ make PREFIX=/export/apps/ray/2.3.1 HAVE_LIBZ=y HAVE_LIBBZ2=y $ strip Ray $ sudo mkdir -p /export/apps/ray/2.3.1 $ sudo chown -R aorth:aorth /export/apps/ray/2.3.1 $ make install $ sudo chown -R root:root /export/apps/ray/2.3.1