mmc2_final.csv

The spreadsheet of indications found in the supplemental materials for this paper ( http://www.sciencedirect.com/science/article/pii/S1359644612002991) has been reformatted in an attempt to make the data more machine readable. <br>First, each unique drug-disease pair has been given its own row in the spreadsheet. To retain the metadata as to whether an indication was "original" or "repurposed", a new column, 'ind_type' has been added with this info. DrugBank IDs and Disease Ontology IDs have been mapped to as many indications as possible. This resulted in a 98% coverage of Drugs and an 85% coverage of diseases (by indication). <br>Indication descriptions have been reduced to the simple Disease Ontology name, or a shortened indication description where no DOID could be mapped. Several were left blank where no mapping could be determined. Indications which are clearly not diseases have been encapsulated in a double asterisk (e.g. **diagnostic imagining** or **analgesic**). The original text of the supplemental spreadsheet has also been retained in the column 'long_ind'.