This tool is a Python implementation of BRAvo. It generates an upstream regulation network from the PathwayCommons knowledge base. pyBRAvo can be used through either a Jupyter notebook, or a command line interface.
A Dockerfile is available to build an image and install the required software environment.
Build the docker container with
docker build -t pybravo .
Run the docker container with
docker run -p 8888:8888 -i pybravo
The first step consists in creating a software environment and pull the required python packages:
conda create --name pybravo python=3.7 rdflib requests matplotlib jupyter networkx flask -c conda-forge -c bioconda
conda activate pybravo
pip install nxpd
Then just clone this repository:
git clone https://github.com/pyBRAvo/pyBRAvo.git
cd pyBRAvo/src
Then test that everything is fine:
python pyBravo.py --regulation --fast --input_genes JUN/FOS SCN5A -md 2 -co -su -sy -excl mirtarbase
You should obtain something like:
Explored 84 regulators
Explored 92 regulators
...
Explored 398 regulators
--- Upstream regulation network in 72.77 seconds ---
Number of nodes = 458
Number of edges = 1683
SIF network written to out.sif
Basic regulation reaction provenance written to out-provenance.csv
| Node | Degree Centrality |
|------|------|
| NOG | 0.162 |
| POU2F1 | 0.144 |
| FOS | 0.144 |
| EGR2 | 0.133 |
| TNF | 0.133 |
| JUN | 0.127 |
| SP1 | 0.125 |
| MAZ | 0.12 |
| LEF1 | 0.114 |
| HNF1A | 0.109 |
--- Network simplification in 0.24 seconds ---
SIF network written to out-unified.sif
Basic regulation reaction provenance written to out-unified-provenance.csv
Nodes after simplification = 436
Edges after simplification = 1657
| Node | Degree Centrality |
|------|------|
| NOG | 0.159 |
| TNF | 0.156 |
| POU2F1 | 0.152 |
| HNF1A | 0.147 |
| FOS | 0.147 |
| EGR2 | 0.14 |
| JUN | 0.136 |
| DAND5 | 0.129 |
| GDNF | 0.129 |
| FGF13 | 0.12 |
For signaling networks :
python pyBravo.py --signaling --input_genes SCN5A -md 2 -co -su -sy -excl mirtarbase --fast
You should obtain something like:
Explored 26 regulators
Explored 3181 regulators
Explored 3197 regulators
Explored 3197 regulators
Explored 3198 regulators
Explored 3198 regulators
Explored 3199 regulators
Explored 3199 regulators
Explored 3199 regulators
--- Upstream regulation network in 95.63 seconds ---
Number of nodes = 3341
Number of edges = 4614
SIF network written to out.sif
Basic regulation reaction provenance written to out-provenance.csv
| Node | Degree Centrality |
|------|------|
| TGF-beta1 | 0.268 |
| HER2 | 0.258 |
| angiotensin II | 0.136 |
| CDK2 | 0.136 |
| DNMT1 | 0.1 |
| L-FoxO1 | 0.096 |
| CLOCK | 0.096 |
| Id | 0.072 |
| LIF | 0.055 |
| Forkhead | 0.013 |
--- Network simplification in 5.39 seconds ---
SIF network written to out-unified.sif
Basic regulation reaction provenance written to out-unified-provenance.csv
Nodes after simplification = 3311
Edges after simplification = 4601
| Node | Degree Centrality |
|------|------|
| TGF-beta1 | 0.269 |
| ERBB2 | 0.26 |
| CDK2 | 0.137 |
| angiotensin II | 0.136 |
| DNMT1 | 0.101 |
| CLOCK | 0.097 |
| L-FoxO1 | 0.096 |
| Id | 0.073 |
| LIF | 0.063 |
| Forkhead | 0.013 |
Inside the src directory, just run the jupyter-notebook BRAvo-python-API-tutorial.ipynb command.
The notebook can be browsed here.
Here is the help message for the command line when running python pyBravo.py:
Please specify one of -reg (--regulation), -sig (--signaling), or -w (--web) option
usage: pyBravo.py [-h] [-w] [-reg] [-sig] [-sigd] [-md MD] [-sy] [-su] [-co] [-fa] [-i I [I ...]] [-f F] [-o O] [-incl INCL [INCL ...]]
[-excl EXCL [EXCL ...]] [-e ENDPOINT] [-unk] [-v]
BRAvo upstream regulation network reconstruction.
Here are some possible command lines :
python pyBravo.py --web
python pyBravo.py --regulation --input_genes JUN/FOS SCN5A -md 2 -co -su -sy
python pyBravo.py --regulation --input_genes JUN/FOS SCN5A -md 2 -excl mirtarbase -co -su -sy
python pyBravo.py --regulation --input_file myGenes.csv -md 2 -incl pid panther msigdb kegg -co -su -sy
Please report any issue to alban.gaignard@univ-nantes.fr.
optional arguments:
-h, --help show this help message and exit
-w, --web to launch pyBravo as a web server
-reg, --regulation to assemble a regulation network
-sig, --signaling to assemble a signaling network
-sigd, --signaling-detailed
to assemble a signaling network with detailed reactions
-md MD, --max_depth MD
the maximum exploration depth
-sy, --extend_with_synonyms
if specified, explore also synonyms
-su, --extend_with_rna_protein_suffixes
if specified, explore also names suffixed with " rna" or " protein"
-co, --decompose_complexes
if specified, decompose protein complexes
-fa, --fast if specified, only explore biopax display names
-i I [I ...], --input_genes I [I ...]
the input gene list
-f F, --input_file F the input file, one gene per line
-o O, --output_file O
the output files path and prefix
-incl INCL [INCL ...], --include_sources INCL [INCL ...]
the data sources to include
-excl EXCL [EXCL ...], --exclude_sources EXCL [EXCL ...]
the data sources to exclude
-e ENDPOINT, --endpoint ENDPOINT
the endpoint to query (default: http://rdf.pathwaycommons.org/sparql/)
-unk, --unknown if specified, do not consider unsigned edges
-v, --verbose print debug information
M Lefebvre, A Gaignard, M Folschette, J Bourdon, C Guziolowski, Large-scale regulatory and signaling network assembly through linked open data, Database, Volume 2021, 2021, baaa113, https://doi.org/10.1093/database/baaa113