Enhancing performance of protein and gene name recognizers with filtering and integration strategies

Abstract

Named entity (NE) recognition is a fundamental task in biological relationship mining. This paper considers protein/gene collocates extracted from biological corpora as restrictions to enhance the precision rate of protein/gene name recognition. In addition, we integrate the results of multiple NE recognizers to improve the recall rates. Yapex and KeX, and ABGene and Idgene are taken as examples of protein and gene name recognizers, respectively. The precision of Yapex increases from 70.90 to 85.84% at the low expense of the recall rate (i.e., it only decreases 2.44%) when collocates are incorporated. When both filtering and integration strategies are employed together, the Yapex-based integration with KeX shows good performance, i.e., the F-score increases by 7.83% compared to the pure Yapex method. The results of gene recognition show the same tendency. The ABGene-based integration with Idgene shows a 10.18% F-score increase compared to the pure ABGene method. These successful methodologies can be easily extended to other name finders in biological documents.

Using UMLS to construct a generalized hierarchical concept-based dictionary of brain functions for information extraction from the fMRI literature

“…IE is a method that allows automatic recognition of meaningful words or phrases from unstructured text. A variety of IE methods have been applied to bioinformatics, either in dictionary-based [5] or rule-based approaches [6,7], in applications including detection of disease, protein and gene names [8][9][10][11]. IE methods have also been used for identifying relationships between different terms-for example, protein-protein interactions [12][13][14][15].…”

Two learning approaches for protein name extraction

“…The Max-Ent method that uses a ''dictionary tagger", achieved the best result among the others with F = 57.9% on the University of Texas, Austin dataset which consists of 748 abstracts. Earlier studies [3,17,[24][25][26][27][28][29][30][31] in the IE community have shown that statistical techniques can be of service in performing protein name extraction tasks. Hidden Markov Models (HMMs), one of the successful statistical learning techniques, have been applied to different IE tasks [3,24,27,30].…”

Constructing Human Brain-Function Association Models from fMRI Literature

“…In recent years, a variety of IE methods have been applied to bioinformatics to detect protein and gene names and construct relationship like protein-protein interaction [3]- [6]. In Wen-Juan et al [7], the goal is to find protein collates extracted from biological corpus to enhance the performance of protein name recognizers. In Barbara et al [8], the task is to classify semantic relations in bioscience text.…”