friguzzi/biokg

A Knowledge Graph for Relational Learning On Biological Data

A knowledge graph for relational learning on biological data.

Compiling BioKG from source

BioKG requires python3.7 or greater to run

Clone the repo

git clone https://github.com/dsi-bdi/biokg.git
cd biokg

Install requirements

pip install -r requirements.txt

BioKG also requires wget. Install it with: Ubuntu:

sudo apt install wget

MacOS:

brew install wget

Compile the BioKG

python run_all.py '<drugbank_username>' '<drugbank_password>'

After the script completes there should be a data folder in the biokg folder This data folder will have 4 folders

• sources which contains the sources used to compile BioKG
• preprocessed which contains the extracted data in preprocessed form
• output which contains the process data and benchmarks
• biokg which contains the final biokg data

There will also be 2 zip files similar to the files contained in the release

• biokg.zip which contains the compressed contents of the biokg folder
• benchmarks.zip which contains the compressed contents of the output/benchmarks folder

Compiling BioKG with docker

Building docker image

sudo docker build . -t dsi-bdi/biokg

Running the image

sudo docker run --rm -v <data_path>:/biokg/data -e DB_USER='<drugbank_username>' -e DB_PASS='<drugbank_password>' dsi-bdi/biokg:latest

• where <data_path> is the fully qualified path to your data folder

Data Sources

The biokg is built using the following data sources.

Source Database	License Type	URL
UniProt	CC BY 4.0	https://www.uniprot.org/help/license
Drugbank	CC BY NC 4.0	https://www.drugbank.ca/legal/terms_of_use
KEGG	Custom	https://www.kegg.jp/kegg/legal.html
Sider	CC BY-NC-SA	http://sideeffects.embl.de/about/
HPA	CC BY SA 3.0	https://www.proteinatlas.org/about/licence
Cellosaurus	CC BY 4.0	https://web.expasy.org/cgi-bin/cellosaurus/faq#Q22
Reactome	CC0	https://reactome.org/license
CTD	Custom	http://ctdbase.org/about/legal.jsp
Intact	Apache 2.0	https://www.ebi.ac.uk/intact/downloads
MedGen	Custom	https://www.nlm.nih.gov/databases/download/terms_and_conditions.html
MESH	Custom	https://www.nlm.nih.gov/databases/download/terms_and_conditions.html
InterPro	Custom	ftp://ftp.ebi.ac.uk/pub/databases/interpro/release_notes.txt
SMPDB	Custom	https://smpdb.ca/about
Hajazi20

Funding

The development of this module has been fully supported by the CLARIFY project that has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 875160.