Table 1. All datasets used in OpenPREDICT version v0.1 and v0.2.

Dataset file	Date retrieved	Data format	Download URL
Bio2RDF r4 datasets (Drugbank, KEGG, HGNC, SIDER and GOA)	2019-08-15	.nq (RDF) compressed as .gz	https://download.bio2rdf.org/#/release/4/
PREDICT drug indication gold standard	2019-08-15	.tab with tabular separator	https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3159979/bin/msb201126-s4.xls
Pubchem-Drugbank mappings	2019-08-15	.tab with tabular separator	https://raw.githubusercontent.com/dhimmel/drugbank/gh-pages/data/mapping/pubchem.tsv
Protein-protein interactions	2019-08-15	.txt with tabular separator	https://science.sciencemag.org/highwire/filestream/628238/field_highwire_adjunct_files/1/Datasets_S1-S4.zip
HPO Phenotype annotations	2019-08-15	.tab with tabular separator	http://compbio.charite.de/jenkins/job/hpo.annotations/lastSuccessfulBuild/artifact/misc/phenotype_annotation.tab
†MESH Phenotype annotations	2019-08-15	.tab with tabular separator	http://www.paccanarolab.org/static_content/disease_similarity/mim2mesh.tsv
MESH Phenotype annotations (BioPortal)	2019-08-15	.txt file	https://raw.githubusercontent.com/fair-workflows/openpredict/master/data/external/meshAnnotationsFromBioPorttalUsingOMIMDesc.txt