|
|
|
|
|
|
|
In the context of the "A Machine-Generated View of the Role of Blood Glucose Levels in the Severity of COVID-19" study, the Blue Brain BioExplorer (BBBE) started as an internal project with the aim to answer key scientific questions related to the Coronavirus as a use case. This project aimed to deliver a visualization tool, the BioExplorer, to reconstruct, visualize, explore and describe in detail the structure and function of the Coronavirus.
Check out the movie by clicking on the following image, and see the coronavirus as you have never seen it before!
The BBBE application is built on top of Blue Brain Brayns, the Blue Brain rendering platform. The BBBE uses the underlying technical capabilities of the rendering platform to create large scale and accurate 3D scenes from Jupyter notebooks.
Assemblies are groups of biological elements, such as proteins, membranes, glycans, etc. As an example, a virion is made of a lipid membrane, spikes proteins, an RNA sequence, etc, and all those elements belong to the same object. That’s why they need to belong to the same container, the assembly. Assemblies can have different shapes: Sphere, Cube, etc, that are automatically generated according to the parameters of individual components.
Proteins are loaded from PDB files. Atoms, non-polymer chemicals and bonds can be loaded and displayed in various colour schemes: chain id, atom, residue, etc. Proteins also contain the amino acid sequences of the individual chains. Sequences that can be used to query glycosylation sites, or functional regions of the protein.
Glycans are small proteins that are attached to an existing protein of the assembly. Individual glycan trees are loaded from PDB files and attached to the glycosylation sites of the specified protein. By default, glycans are attached to all available glycosylation sites, but a set of specific sites can be specified.
Glycan trees models located in the python sdk test folder were generated with Glycam Builder.
An RNA sequence can be loaded from a text sequence of codons. Various shapes can be selected to represent the RNA sequence: Trefoil knot, torus, star, etc. This allows the sequence to be efficiently packed into a given volume. A different color is assigned per type of codon.
Mesh-based membranes create membranes based on 3D meshes. This allows the construction of complex membranes where mesh faces are filled with proteins.
A viral particle (= “virus”) is an assembly consisting of a membrane, an RNA sequence, and a given number of S, M and E proteins. The virus has a predefined spherical shape defined by its radius. The default parameters for the virus are a radius of 45 nanometers, 62 S proteins, 42 E proteins, and 50 M proteins. Dimensions and concentrations were retrieved from the literature.
A membrane is an assembly of phospholipids. Phospholipids structures are created following the process described in the VMD Membrane Proteins tutorial. The assembly itself is generated by the BioExplorer, for a given shape, and a number of instances of phospholipids.
A simple API if exposed via the BBBE python library. The API allows scientists to easily create and modify assemblies, according the biological parameters. The BBBE programming language is not necessarily reflecting the underlying implementation, but is meant to be as simple as close as possible to the language used by the scientists to describe biological assemblies.
The BBBE Python SDK is available on pypi.
A large number of examples (as python notebooks) are provided in the notebooks folder.
See here for detailed documentation of the source code.
BBBE binaries are publicaly available as docker images. BBBE is designed to run in distributed mode, and is composed of 3 modules: A server, a python SDK, and a web user interface. This means that there are 3 docker images to be downloaded on run. Those images can of course run on different machines.
In this example, we will expose the server on port 5000, the python SDK jupyter notebooks on port 5001, and the user inferface on port 5002. One is free to change those ports at will.
docker run -ti --rm -p 5000:8200 bluebrain/bioexplorerdocker run -ti --rm -p 5001:8888 bluebrain/bioexplorer-python-sdkdocker run -ti --rm -p 5002:8080 bluebrain/bioexplorer-ui
In order to run the BioExplorer, it is necessary to build Blue Brain Brayns first. Please refer to the documentation and following the steps in the "Building from source" paragraph. Note that the BioExplorer is currently only supported with version 1.0.0 (e12fa84) of Blue Brain Brayns.
With Blue Brain Brayns compiled and installed in the <brayns_installation_folder>, run the statements to build the BioExplorer.
git clone https://github.com/BlueBrain/BioExplorer.git
mkdir build
cd build
CMAKE_PREFIX_PATH=<brayns_installation_folder> cmake .. -DCMAKE_INSTALL_PREFIX=<brayns_installation_folder>
make installThe BioExplorer being a plug-in for Blue Brain Brayns, the following commands can be used to start the backend:
export LD_LIBRARY_PATH=<brayns_installation_folder>/lib:${LD_LIBRARY_PATH}
export PATH=<brayns_installation_folder>/bin:${PATH}
braynsService --http-server localhost:5000 --plugin BioExplorerConsidering that the BBBE server is running on the local host, on port 5000, the simplest example to visualize a coronavirus is:
from bioexplorer import BioExplorer
be = BioExplorer('localhost:5000')
resource_folder = '../../tests/test_files/'
name='Coronavirus'
be.add_coronavirus(name=name, resource_folder=resource_folder)BBBE is available to download and use under the GNU General Public License (GPL, or “free software”). The code is open sourced with approval from the open sourcing committee and principal coordinators of the Blue Brain Project in February 2021.
For more information on BBBE, please contact:
Cyrille Favreau
Senior Scientific Visualization Engineer
Blue Brain Project
cyrille.favreau@epfl.ch
The development of this software was supported by funding to the Blue Brain Project, a research center of the École polytechnique fédérale de Lausanne (EPFL), from the Swiss government’s ETH Board of the Swiss Federal Institutes of Technology.
COPYRIGHT 2020–2021, Blue Brain Project/EPFL