Datasets used in DrugMint
|
Main Dataset: This dataset contain FDA approved drugs and the molecules that have not apporved yet. The data used in our study was kindly provided by Authors's of Tang K. et. al. 2011. This dataset comprise`s of 1348 approved and 3206 experimental drugs compiled from DrugBank2.5. Our main dataset contain 1347 instead 1348 as we were not able to compute descriptor of one molecule "Teicoplanin"..
Independent Dataset: The independent dataset was created by extracting molecules from the DrugBank3.0 database, that are not present in DrugBank2.5. This dataset contains 100 approved drugs and 1964 experimental drug molecules. This dataset was used to evaluate the performance of our prediction method. Derived Dataset: The derived dataset was extracted by comparing DrugBank3.0 and its previous version DrugBank2.5. The drugs that were running in experimental phage of drug discovery during the compilation of DrugBank2.5, some of them were get approved for human use before the release of DrugBank3.0. These 21 drugs were very useful to evaluate performance of our model. As per expectation these drugs were predicted in Drug-like category by Drugmint Server. User can download different datasets used in this study from following table.
|