Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data
ICML '01 Proceedings of the Eighteenth International Conference on Machine Learning
A simple rule-based part of speech tagger
HLT '91 Proceedings of the workshop on Speech and Natural Language
Detecting duplicate biological entities using Shortest Path Edit Distance
International Journal of Data Mining and Bioinformatics
| Hi-index | 0.00 |
The side effect of drugs often results from a response to the unintended target of a drug. Recently there have been researches identifying targets of known drugs based on the side effect information. These researches, however, did not consider the association of drugs both with targets and with biological processes. The recent development of a database of the side effects, SIDER, is a first step to provide the relationship between drugs and side effects. In this paper, we propose a novel process-drug-side effect network that discovers the relationship between biological processes and side effects. The multi-level network (the process-drug-side effect network) is built by merging the drug-biological process network and the drug-side effect network. Evaluation is conducted in two ways: 1) how many biological processes discovered by our method are found in co-occurred GO terms with biological processes extracted from the PubMed records by a text mining technique. 2) whether there is the performance improvement by limiting response processes by drugs sharing the same side effect only to frequent ones. The experimental results show that our process-drug-side effect network is able to discover meaningful relationships between biological processes and side effects in an efficient manner.