EGenPub, a text mining system for extending computationally mapped bibliography for UniProt Knowledgebase by capturing centrality.

ABSTRACT: UniProt Knowledgebase (UniProtKB) is a publicly available database with access to a vast amount of protein sequence and functional information. To widen the scope of the publications associated with a protein entry, UniProt has introduced the computationally mapped additional bibliography section, which includes literature collected from external sources. In this article, we describe a text mining system, eGenPub, which selects articles that are 'about' specific proteins and allows automatic identification of additional bibliography for given UniProt protein entries. Focusing on plant proteins initially, eGenPub utilizes a gene normalization tool called pGenN, and a trained support vector machine model, which achieves a precision of 95.3%, to predict whether an article, based on its abstra

SUBMITTER: Ding R 

PROVIDER: S-EPMC5691349 | biostudies-literature | 2017 Jan

REPOSITORIES: biostudies-literature