APORC Document Center: NetPredATC

More...

NetPredATC: Network predicting drug's anatomical therapeutic chemical code

Version 0.0.1
Last updated: Oct. 16, 2012

References#

Yong-Cui Wang, Nai-Yang Deng, and Yong Wang. Network predicting drug's anatomical therapeutic chemical code. Bioinformatics, (2013) doi: 10.1093/bioinformatics/btt158 PDF

Motivation#

Discovering drug's Anatomical Therapeutic Chemical (ATC) classification rules at molecular level is of vital importance to understand a vast majority of drugs action. However, few studies attempt to annotate drug's potential ATC-codes by computational approaches.

Method#

Here, we introduce the drug-target network and data integration to computationally predict drug's ATC-codes and propose the NetPredATC method. Starting from the assumption that drugs with similar chemical structures or target proteins share common ATC-codes, we propose a novel method to assign drug's potential ATC-codes by integrating chemical structures and target proteins. Specially, we first construct a gold-standard positive dataset for drug and ATC-code interactions by knowledge. Then, we characterize ATC-code and drug by their similarity-based profiles, and define the kernel function to correlate drug with ATC-code. Finally, we train a machine learning model, support vector machine (SVM), to automatically predict drug's ATC-codes. Our method is validated on four drug datasets with various target proteins, including enzymes, ion channels (ICs), G-protein couple receptors (GPCRs), and nuclear receptors (NRs). We find that both drug's chemical structure and target protein are predictive and target protein information has better accuracy. Integrating these two features, our method outperforms the chemical similarity only based method and more experimentally validated ATC-codes for drugs can be revealed. In addition, database search and functional annotation analysis indicate that our new predictions are worthy of future experimental validation.