Add results

README.md

@@ -1,2 +1,10 @@
-# results
+# Results
 Applications and validation analyses shown in the manuscript
+
+## Networks
+The two network files used in the case scenario.
+
+## Validation
+To validate the disease-drug pairs prioritized by Drug2ways, we used clinical trial information (i.e., drugs that have
+been tested on a given indication) from [ClinicalTrials.gov](https://clinicaltrials.gov/) mapped to each of the networks.
+

networks/README.md

@@ -0,0 +1,56 @@
+# Drug2Ways Networks
+
+This folder contains networks used in the Drug2Ways application scenarios. These networks include:
+
+* An In-House network
+* OpenBioLink knowledge graph
+
+## In-House Network
+
+The In-House network contains causal interactions (i.e., -1 for decrease and +1 for increase) between the
+ following entity types:
+
+* **drug-gene**
+* **gene-gene**
+* **gene-disease**
+
+Resources used to create this network include the following:
+* **drug-gene** interactions were extracted exclusively from [DrugBank](https://www.drugbank.ca/)
+* **gene-gene** interactions were extracted from:
+    * [BioGRID](https://thebiogrid.org/)
+    * [IntAct](https://www.ebi.ac.uk/intact/)
+    * [KEGG](https://www.genome.jp/kegg/)
+    * [PathwayCommons](https://www.pathwaycommons.org/)
+    * [Reactome](https://reactome.org/)
+    * [WikiPathways](https://www.wikipathways.org/)
+* **gene-disease** interactions were extracted exclusively from [DisGeNET](https://www.disgenet.org/)
+
+## OpenBioLink Knowledge Graph
+
+The [OpenBioLink](https://github.com/OpenBioLink/OpenBioLink) knowledge graph (KG) was used to generate a second network
+ used for the Drug2Ways case scenario. The OpenBioLink benchmark dataset is a heterogeneous network incorporating a
+ diverse set of biological entities and relations. Causal relationships in the KG have been inferred and simplified to
+ -1 for decrease and +1 for increase relationships. 
+
+## Directories
+
+#### Create Networks
+In the create_networks directory, you can find notebooks and scripts to generate the In-House pharmacological network or
+ the modified OpenBioLink network.
+
+There you can find the following resources:
+
+* The create_custom_network.ipynb generates the In-House pharmacological network, retaining only causal relationships.
+* The map_relations.py file contains methods to download the OpenBioLink KG and map interactions between a subset of
+ node types to BEL relations.
+
+#### Data 
+In the data directory, you can find TSV files containing the interactions found in the In-House and modified OpenBioLink
+networks.
+
+There you can find the following resources:
+
+* The custom_network.tsv file contains causal interactions between genes, drugs and diseases derived from the resources
+ described above.
+* The openbiolink_network.csv file contains causal interactions between genes, drugs, diseases and phenotypes derived
+ from the OpenBioLink KG.