StreaMD: a tool to perform high-throughput automated molecular dynamics simulations

This is Apptainer (Singularity-compatible) container images providing a ready-to-run installation of StreaMD (CPU/GPU), a toolkit for high-throughput molecular dynamics (MD) simulations that automates preparation, equilibration, production, and analysis workflows for protein-only, protein–ligand, and protein–cofactor systems. 

Project resources

• Source code (GitHub): ci-lab-cz/streamd. GitHub

• Documentation (Read the Docs): StreaMD Documentation (overview, installation, CLI reference, config file, GBSA/MM-PBSA, ProLIF, etc.).

Both streamd_cpu.sif and streamd_gpu.sif are built from the same StreaMD codebase (streamd version 0.4) and expose the same command-line interfaces (run_md, run_gbsa, run_prolif). All CLI options and parameters are identical between github version and CPU/GPU images; the only difference is runtime execution on GPU nodes, which requires --nv for the GPU image.

streamd_cpu.sif (CPU container) 

This is an Apptainer CPU container image for StreaMD intended for execution on CPU nodes.

Project resources

• Documentation (Read the Docs): StreaMD Documentation/cpu-usage.

Usage

Example:

Advanced analysis tools:

StreaMD supports YAML configuration via --config (command-line arguments override config values). StreaMD Documentation/configuration

There is an option `-c/--ncpu` to control the number of CPUs used. By default all CPUs will be used.
Example:

streamd_gpu.sif (GPU container)

This is an Apptainer GPU container image for StreaMD intended for execution on NVIDIA GPU nodes. GPU runs should be launched with --nv to expose NVIDIA devices and driver libraries to the container.

Project resources

• Documentation (Read the Docs): includes a dedicated section on GPU usage and runtime behavior. StreaMD Documentation/parallelism-and-gpus

Usage on GPU nodes

Example: