Finding relevant protein of ADRs using text mining in Taverna

Abstract

Adverse Drug Reactions (ADRs) are the undesirable reactions from use of medicinal products effecting protein through drugs or their reactive metabolites binding. These are the fourth leading cause of death in the US. More than nineteen drugs have been withdrawn since 1998 because of their adverse reactions. These incidents diminish the medical confidences and value of pharmaceutical industry. Many attempts have been made to explicate the relevance of proteins and ADRs which reveal their underlying biological mechanisms. These will be useful for adverse reactions prediction. However, the rapid growth in published scientific literatures leads to the difficulty of being up to date and figuring out the relevant data for researchers. Text mining is a solution for this trouble by extracting information from written resources. By implementing text mining and Taverna, the text mining function is presented in a single graphical workflow. There are 2,551 proteins that discovered to associated with 83 Preferred Terms of hepatobiliary disorders. The collection of these proteins and their relavence then will be connected with ADRs and drug information. The relations might be used for facilitating better understanding of the molecular mechanisms of drug-induced ADRs and the rational drug discovery.