Boya Xie1, Qin Ding, Hongjin Han, Di Wu. 1. Department of Computer Science, East Carolina University, Greenville, NC 27858 and Department of Physiology, Brody School of Medicine, East Carolina University, Greenville, NC 27834, USA.
Abstract
MOTIVATION: Research interests in microRNAs have increased rapidly in the past decade. Many studies have showed that microRNAs have close relationships with various human cancers, and they potentially could be used as cancer indicators in diagnosis or as a suppressor for treatment purposes. There are several databases that contain microRNA-cancer associations predicted by computational methods but few from empirical results. Despite the fact that abundant experiments investigating microRNA expressions in cancer cells have been carried out, the results have remain scattered in the literature. We propose to extract microRNA-cancer associations by text mining and store them in a database called miRCancer. RESULTS: The text mining is based on 75 rules we have constructed, which represent the common sentence structures typically used to state microRNA expressions in cancers. The microRNA-cancer association database, miRCancer, is updated regularly by running the text mining algorithm against PubMed. All miRNA-cancer associations are confirmed manually after automatic extraction. miRCancer currently documents 878 relationships between 236 microRNAs and 79 human cancers through the processing of >26 000 published articles. AVAILABILITY: miRCancer is freely available on the web at http://mircancer.ecu.edu/
MOTIVATION: Research interests in microRNAs have increased rapidly in the past decade. Many studies have showed that microRNAs have close relationships with various humancancers, and they potentially could be used as cancer indicators in diagnosis or as a suppressor for treatment purposes. There are several databases that contain microRNA-cancer associations predicted by computational methods but few from empirical results. Despite the fact that abundant experiments investigating microRNA expressions in cancer cells have been carried out, the results have remain scattered in the literature. We propose to extract microRNA-cancer associations by text mining and store them in a database called miRCancer. RESULTS: The text mining is based on 75 rules we have constructed, which represent the common sentence structures typically used to state microRNA expressions in cancers. The microRNA-cancer association database, miRCancer, is updated regularly by running the text mining algorithm against PubMed. All miRNA-cancer associations are confirmed manually after automatic extraction. miRCancer currently documents 878 relationships between 236 microRNAs and 79 humancancers through the processing of >26 000 published articles. AVAILABILITY: miRCancer is freely available on the web at http://mircancer.ecu.edu/
Authors: Eman A Toraih; Manal S Fawzy; Eman A Mohammed; Mohammad H Hussein; Mohamad M El-Labban Journal: Mol Diagn Ther Date: 2016-12 Impact factor: 4.074