For a pilot study in medical research, define a method to predict the pharmaceutical potential of new active molecules.
An artificial neuronal network was created with the respective input parameters for the data preparation and transformation for the input layers. The network was automatically trained with sample data and the abstraction capability was verified with test data.
The new, trained network was able to predict the potential therapeutic chemical classification system (ATC) with an accuracy of 70%. It is able to support decision making in the very first phase of drug development in the pre-clinical research of new molecules.