README_for_docker.md

MEGADOCK with Docker

Docker container images are available on akiyamalab/megadock repository (DockerHub).

• akiyamalab/megadock:cpu
• akiyamalab/megadock:gpu

Requirements

Requirement Tools	GPU	CPU	Notes
Docker	x	x
nvidia-docker	x		> 2.0

Quick Example

# CPU single node (OpenMP parallelization)
docker run akiyamalab/megadock:cpu megadock -R data/1gcq_r.pdb -L data/1gcq_l.pdb -o data/1gcq_r-1gcq_r.out

# GPU single node (GPU parallelization)
docker run --runtime=nvidia akiyamalab/megadock:gpu megadock-gpu -R data/1gcq_r.pdb -L data/1gcq_l.pdb -o data/1gcq_r-1gcq_r.out

---

Build Docker Container for MEGADOCK (CPU)

1. build Docker image

# on ${REPOSITORY_ROOT} dir
docker build . -f Dockerfiles/cpu/Dockerfile -t akiyamalab/megadock:cpu

2. run sample

docker run akiyamalab/megadock:cpu megadock -R data/1gcq_r.pdb -L data/1gcq_l.pdb

# optional) start interactive shell
docker run -it akiyamalab/megadock:cpu

# optional) run with your pdb (e.g. ${DATA_PATH} = your pdb-data directory abs path  )
docker run -v ${DATA_PATH}:/opt/MEGADOCK/data akiyamalab/megadock:cpu megadock -R data/${RECEPTOR}.pdb -L data/${LIGAND}.pdb

Build Docker Container for MEGADOCK (GPU)

nvidia-docker is required.

1. build Docker image

# on ${REPOSITORY_ROOT} dir
docker build . -f Dockerfiles/gpu/Dockerfile -t akiyamalab/megadock:gpu

2. run sample

docker run -it --runtime=nvidia akiyamalab/megadock:gpu megadock-gpu -R data/1gcq_r.pdb -L data/1gcq_l.pdb

# optional) start interactive shell
docker run -it --runtime=nvidia akiyamalab/megadock:gpu

# optional) run with your pdb (e.g. ${DATA_PATH} = your pdb-data directory abs path  )
docker run -it --runtime=nvidia -v ${DATA_PATH}:/opt/MEGADOCK/data akiyamalab/megadock:gpu megadock-gpu -R data/${RECEPTOR}.pdb -L data/${LIGAND}.pdb