https://github.com/becksteinlab/conda-gromacs-2016
Conda package of gromacs 2016 intended for testing purposes.
https://github.com/becksteinlab/conda-gromacs-2016
conda gromacs
Last synced: 3 months ago
JSON representation
Conda package of gromacs 2016 intended for testing purposes.
- Host: GitHub
- URL: https://github.com/becksteinlab/conda-gromacs-2016
- Owner: Becksteinlab
- License: mit
- Created: 2016-09-21T23:48:53.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2017-02-28T00:18:33.000Z (about 8 years ago)
- Last Synced: 2025-01-06T10:48:27.997Z (5 months ago)
- Topics: conda, gromacs
- Language: Shell
- Homepage:
- Size: 5.86 KB
- Stars: 0
- Watchers: 3
- Forks: 0
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# README for conda-gromacs-2016
This repository contains everything needed to build
[conda](https://conda.io/docs/) packages for the popular
[Gromacs](http://www.gromacs.org) molecular dynamics package.
**WARNING**: The package builds a **minimal, portable Gromacs version** that is
primarily useful for testing. It is **not recommended to use it for
production-scale simulations.**
Reduced performance features:
* only thread parallelization
* uses bundled FFT (not a fast FFT)
* no GPU support
## Using the installed package
Multiple versions of Gromacs can be installed in parallel. A script
`get_gmx` is installed in the user's home directory that allows the
user to select the desired version of Gromacs. It returns the path to
the `GMXRC` startup script, which in turn needs to be sourced to set
up the Gromacs environment. For instance, to select version
*2016* one would run in the shell
```
source `get_gmx 2016`
```
## Pre-built packages
Anaconda packages for *linux-64* are available in the Anaconda Cloud channel
[anaconda.org/becksteinlab](https://anaconda.org/becksteinlab). If you
have `conda` installed you can install this package with
```
conda install -c becksteinlab gromacs-2016
```
## Building the package
It is recommended to build the package within a minimal docker container.
Build the docker container by invoking
```
docker build -t gmx_build .
```
in the same directory as the provided Dockerfile.
Start the docker container with `docker run -it gmx_build`. If everything
works as expected, then you should find yourself in an environment similar
to:
```
[root@0f6a75177395 /]#
```
Navigate to `/root/gromacs` to find the build scripts. Run:
```
bash build_conda_package.sh
```
Once this has finished move to the parent directory to find your package in tar.bz2
archive.
### Keep locally or push to cloud:
#### Copy to host machine
From a new terminal find the container ID with `docker ps`. With this ID, invoke:
```
docker cp :/root/ $OUTPUT_DIR
```
You should be able to find the conda package in the output directory.
To exit the docker container type `exit`.
#### Upload directly to Anaconda cloud
To upload directly to a the Anaconda clound through your Anaconda username, run:
```
anaconda upload -u
```
To exit the docker container type `exit`
## Licensing
[Gromacs](http://www.gromacs.org/About_Gromacs) itself is published
under the GNU
[Lesser General Public License](http://www.gnu.org/licenses/lgpl-2.1.html)
(LGPL), version 2.1. The files to *build anaconda packages* are
licensed under the MIT license (see file `LICENSE`).
## Contributing
We welcome contributions in the form of issue reports and pull requests.