Drug Target Negative Interactions Dataset
The dataset represents the negative interaction dataset of the Drugbank that has been generated from our proposed machine learning method based on drug similarity, which achieved an average accuracy of 95% compared to the randomly generated negative datasets in the literature. Drugbank was used as the drug target interaction dataset from https://go.drugbank.com/. It consists of 1,264 interactions among 504 drugs and 507 proteins. The dataset includes drugs names, their accession numbers, proteins names, their UniproteId on Uniprot at https://www.uniprot.org/. Drug features were obtained to get proteins sequences or extraction of protein features from Drugbank.
The dataset consists of one excel file, representing the Negative interaction file, and consist of two columns:
- The first column is the Drug id for each drug in Drugbank.
- The second column is a list negative interaction proteins and each element representing the name of protein in Uniprot.