Hello,
I already made a ligand-based pharmacophore by utilizing 60 dataset compounds by catalyst module in discovery studio. But I want to explore the ROC curve of that pharmacophore in order to find true predictability of that pharmacophore.
Can anyone guide me in the discovery studio?
You need to perform the validation step and for that you must have two datasets of actives and inactives.
Anu Manhas In 3d qsar ligand-based pharmacophore modeling module, there is a validation option where only Fischer randomization and test set validation option is there. So, there is no option about active and inactive test set incorporation. Only active and inactive test set incorporation is under common feature pharmacophore generation module. So, how can I perform this?
Common feature pharmacophore generation involves the generation of the pharamacophore from ligands only. So, you can perform the additional validation procedure you are asking about.
But I was doing activity based pharmacophore generation. If i do common feature pharmacophore generation that time the resultant pharmacophore was different. So what should i do?
- Badges
- Science topic