aniewielska / biobb_workflows

BioExcel building blocks tutorials

biobb_workflows

The Jupyter notebook Lysozyme.ipynb shows a Lysozyme + Mutations tutorial using the BioExcel building blocks biobb.

You may execute the notebook on a free cloud instance using myBinder, or for better performance/extensibility in your on instance of Jupyter Notebook.

Docker image

The Docker image bioexcel/biobb_workflows is provided, based on myBinder.

Launch it as:

docker run -p 8888:8888 -it bioexcel/biobb_workflows

Then access the Notebook at http://127.0.0.1:8888/ using the provided token.

Tip: If you have an NVIDIA GPU and NVIDIA Docker installed, launch with nvidia-docker instead of docker for hardware-accelerated GROMACS.

BioConda installation

For your own installation it is recommended to use BioConda on Linux or OS X. See the BioConda installation instructions for details. In brief:

First, install MiniConda (Python 3.x recommended). Then activate BioConda channels:

conda config --add channels defaults
conda config --add channels bioconda
conda config --add channels conda-forge

Make a new Conda environment to keep our dependencies:

conda create -n biobb biobb_md biobb_io biobb_model nglview=1.2.0 simpletraj gromacs

Tip: Remove gromacs above if you have already installed GROMACS that you compiled with hardware accelerations like NVidia CUDA.

Activate the Conda environment:

conda activate biobb

OpenCL implementations

After activating the Conda environment for the first time you may see a warning equivalent to:

WARNING: No ICDs were found. Either,
- Install a conda package providing a OpenCL implementation (pocl, oclgrind) or 
- Make your system-wide implementation visible by doing 'ln -s /etc/OpenCL/vendors /home/USER/miniconda3/envs/biobb/etc/OpenCL/vendors'  

The PyOpenCl documentation suggests how to proceed. OS X users should use conda install khronos-opencl-icd-loader which can use GPU hardware installed, while in Linux you can either provide a software-based OpenCL implementation with conda install pocl or (recommended) use GPU-specific hardware implementation found in /etc/OpenCL/vendors (e.g. apt install nvidia-driver-390) by creating the symlink as suggested.

Start Jupyter Notebook

Activate the nglview extension and start Jupyter Notebook:

jupyter-nbextension enable nglview --py --sys-prefix
jupyter notebook biobb_workflows/lysozyme_tutorial/notebooks/Lysozyme.ipynb

A browser window should appear, or click the link provided.

Copyright & Licensing

This software has been developed in the MMB group (http://mmb.irbbarcelona.org) at the BSC (http://www.bsc.es/) & IRB (https://www.irbbarcelona.org/) and eScience Lab group at The University of Manchester for the European BioExcel (http://bioexcel.eu/) CoE, funded by the European Commission (EU H2020 675728, 823830).

• (c) 2018-2019 Barcelona Supercomputing Center
• (c) 2018-2019 Institute for Research in Biomedicine
• (c) 2018-2019 The University of Manchester

The Jupyter Notebook is based on the Lysozyme in Water GROMACS tutorial by Justin A. Lemkul at Virginia Tech Department of Biochemistry.

Licensed under the Apache License 2.0, see the file LICENSE for details.