Background: Computer-aided data mining methods and sophisticated softwares are used to examine candidate molecules during the drug discovery process. Data mining and machine learning are effective tools in leveraging the drug datasets.
Objective: Developed model in this study can be used as a simple filter in drug modelling to eliminate potentially inappropriate molecules in the early stages. In this work we developed a Drug Decision Support System (DDSS), in which these tools are used to induce classification models, association rules and subgraph patterns. DDSS helps drug designers to develop a successful drug candidate.
Methods: Molecular descriptors that are effective in classification models were identified for determination of a number of rules in drug molecules. They are derived using ADRIANA.Code program and Lipinski's rule of five. Closed frequent molecular structures in the form of subgraph fragments were also obtained with Gaston algorithm. Gaston algorithm, included in ParMol Package (Parallel Molecular Mining) was used to find common molecular fragments in the drug datasets. WEKA machine learning tool version 3.6.11 and MATLAB software package (MATLAB & SIMULINK, R2015a) were used as tools for this study.
Results: We observed that TPSA, XlogP Natoms, HDon_O and TPSA are the most distinctive features in the pool of the molecular descriptors. Cardiac therapy, anti-epileptics and anti-parkinson drugs with approved and withdrawn drugs are identified and related databases are screened to obtain datasets for experimentation.
Conclusion: The experimental evaluation shows that the system is promising at determination of potential drug molecules to classify drug molecules correctly according to the types of diseases.