This repository contains exports of Hetionet as Biological Expression Language (BEL). The research article describing Project Rephetio and Hetionet v1.0 is:

Systematic integration of biomedical knowledge prioritizes drugs for repurposing Daniel S Himmelstein, Antoine Lizee, Christine Hessler, Leo Brueggeman, Sabrina L Chen, Dexter Hadley, Ari Green, Pouya Khankhanian, Sergio E Baranzini eLife (2017-09-22) DOI: 10.7554/eLife.26726

BEL is a domain specific language that enables the expression of biological relationships in a machine-readable format. It is supported by the PyBEL software ecosystem.

The network is available in three BEL formats:

• BEL Script - see description below
• Nodelink JSON - see description below
• GraphDati JSON - see description below

Large files in this repository are stored using Git LFS. When cloning this repository, please make sure that Git LFS is installed on your system. Otherwise, git will checkout text pointers for large files rather than the large files themselves!

The table below presents the way each metaedge from Hetionet is mapped to BEL, with a few caveats and notes:

• In the case of Compound - binds - Gene (CbG), the information about the pharmacological action was used to split this edge into several more types.
• The metaedges Compound - palliates - Disease (CpD) and Compound - treats - Disease (CtD) are collapsed into the same BEL edge a(X) decreases path(Y)
• The metaedges Disease - resembles - Disease (DrD) and Disease - presents - Symptom (DpS) look the same because diseases and symptoms are both represented by the path() BEL function, but they differ in namespaces. Future BEL updates might allow for finer granularity nomenclature of entity types beyond path().
• The metaedges Gene - participates - Biological Process (GpBP) and Gene - participates - Pathway (GpPW) look the same because biological processes and pathways are both represented with the bp() BEL function.
• Gene Ontology molecular activities and cellular components can't easily be represented in BEL, so the Gene - participates - Cellular Component (GpCC) and Gene - participates - Molecular Function (GpMF) edges are removed
• The metaedges Compound - downregulates - Gene (CdG) and Compound - upregulates - Gene (CuG) are both considered as being related to the amount of the protein. This assumption could be wrong (did some of these data come from LINCS microarray expreriments, for example? In that case it should be abundance of the RNA)

This repository redistributes content from hetio/hetionet and is licensed in the same way. See the License section of the original Hetionet repository and moore information on Thinklab.

BEL Script is the de facto standard for BEL, which all BEL-aware applications should be able to consume. It contains informations about the nodes, edges, and their biological context in a domain-specific language. It can be parsed with PyBEL or other BEL parsers.

Example opening BEL Script using pybel.from_bel_script_gz():

from pybel import from_bel_script_gz
graph = from_bel_script_gz('hetionet-v1.0.bel.gz')

Node-link is the format popularized by Javascript frameworks like D3 for representing network information. Since the main data structire in PyBEL is a network, it often makes sense to use Nodelink JSON as a pre-compiled data structure for BEL (since parsing/compiling BEL takes a lot longer than JSON). The schema is specific to PyBEL, but this is the fastest to load.

Example opening Nodelink JSON using pybel.from_nodelink_gz():

from pybel import from_nodelink_gz
graph = from_nodelink_gz('hetionet-v1.0.bel.nodelink.json.gz')

GraphDati JSON is another JSON schema used for BEL by the BEL.bio and BioDati projects (note: username/password for the demo server are demo/demo).

In general, BEL graphs can be exported to GraphDati JSON then uploaded to BioDati via its API. Note, this address will be different for your instance of BioDati. More directly, BEL graphs in PyBEL can be uploaded programatically with pybel.post_graphdati().