Wolf parameter estimation along the vapor-liquid coexistence curve
Running this on the WSU grid environment requires pulling singularity containers before running the workflow. Make a directory, accessible from compute nodes, for storing the images $ mkdir ~/singularity_cache Some grid environments, WSU for example, require using the module to build images. $ module load singularity Pull the singularity images individually $ singularity pull --name go2432-gomc-cpu.img docker://go2432/gomc:cpu $ singularity pull --name go2432-namd-latest.img docker://go2432/namd:latest $ singularity pull --name go2432-mosdef-gomc-latest.img docker://go2432/mosdef-gomc:latest $ singularity pull --name go2432-scikit-optimize-latest.img docker://go2432/scikit-optimize:latest $ cd ~/WolfWater The singularity.conf file has to match that of the grid environment in the conda environment. $ cp /opt/ohpc/pub/libs/singularity/3.4.1/etc/singularity/singularity.conf /wsu/home/go/go24/go2432/mambaforge/envs/nextflow/etc/singularity A working running script to launch the nextflow pipeline is included. It sets some environment variables, activates a mamba environment with nextflow and singularity, and runs the pipeline. $ sbatch headless.sh