Using cellfinder at the SWC
The compute cluster at the SWC has BrainGlobe tools preinstalled, and so the instructions vary slightly from that in the Installation with SLURM section.
Before starting to use cellfinder on the SWC. It is recommended to familarise yourself with the job scheduler system (SLURM).
On the SWC cluster, no software needs to be installed, as cellfinder can be loaded with
module load brainglobe.Cellfinder can be used interactively, by starting an interactive job:
srun -p gpu --gres=gpu:1 -n 20 -t 0-24:00 --pty --mem=120G bash -i
Loading cellfinder:
module load brainglobe
Loading CUDA:
module load cuda/11.2
It is recommended to use cellfinder by using the batch submission system. This has many advantages:
- Your analysis is reproducible (you have a script showing exactly what you did)
- You don't need to wait for computing resources to become available (once submitted, the job will wait until it can be run)
- If for any reason the analysis is interrupted, you can easily restart
- You don't need to keep a connection to the cluster open
- You can easily receive email updates when the job starts and finishes.
An example batch script is given below, but it is recommended to familarise yourself with the batch submission system before trying to optimise cellfinder.
#!/bin/bash
#SBATCH -p gpu # partition (queue)
#SBATCH -N 1 # number of nodes
#SBATCH --mem 120G # memory pool for all cores
#SBATCH --gres=gpu:1
#SBATCH -n 20
#SBATCH -t 2-0:0 # time (D-HH:MM)
#SBATCH -o cellfinder.out
#SBATCH -e cellfinder.err
#SBATCH --mail-type=ALL
#SBATCH [email protected]
cell_file='/path/to/signal/channel'
background_file='path/to/background/channel'
output_dir='/path/to/output/directory'
echo "Loading cellfinder environment"
module load brainglobe
module load cuda/11.2
echo "Running cellfinder"
# Just an example. See the user guide for the specific parameters
cellfinder -s $cell_file -b $background_file -o $output_dir -v 5 2 2 --orientation psl
Last modified 7mo ago