BEST: Next-Generation Biomedical Entity Search Tool for Knowledge Discovery from Biomedical Literature

Lee, Sunwon; Kim, Donghyeon; Lee, Kyubum; Choi, Jaehoon; Kim, Seongsoon; Jeon, Minji; Lim, Sangrak; Choi, Donghee; Kim, Sunkyu; Tan, Aik Choon; Kang, Jaewoo

doi:10.1371/journal.pone.0164680

Cited by 90 publications

(75 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Computational procedure for identifying fusion-related PubMed abstracts is summarized in Figure 1 with more detailed information provided in Supplementary Figure S1. Out of ∼26 million PubMed abstracts as of June 2016, we searched all sentences containing multiple gene names that were recognized by the BEST entity extractor (19) after taxonomy filtering to remove articles on non-human species using PubTator (20). This initial screening yielded 302 615 sentences in 156 229 abstracts.…”

Section: System Design and Methodsmentioning

confidence: 99%

OUP accepted manuscript

Lee

et al. 2016

Nucleic Acids Research

View full text Add to dashboard Cite

Fusion gene is an important class of therapeutic targets and prognostic markers in cancer. ChimerDB is a comprehensive database of fusion genes encompassing analysis of deep sequencing data and manual curations. In this update, the database coverage was enhanced considerably by adding two new modules of The Cancer Genome Atlas (TCGA) RNA-Seq analysis and PubMed abstract mining. ChimerDB 3.0 is composed of three modules of ChimerKB, ChimerPub and ChimerSeq. ChimerKB represents a knowledgebase including 1066 fusion genes with manual curation that were compiled from public resources of fusion genes with experimental evidences. ChimerPub includes 2767 fusion genes obtained from text mining of PubMed abstracts. ChimerSeq module is designed to archive the fusion candidates from deep sequencing data. Importantly, we have analyzed RNA-Seq data of the TCGA project covering 4569 patients in 23 cancer types using two reliable programs of FusionScan and TopHat-Fusion. The new user interface supports diverse search options and graphic representation of fusion gene structure. ChimerDB 3.0 is available at http://ercsb.ewha.ac.kr/fusiongene/.

show abstract

Section: System Design and Methodsmentioning

confidence: 99%

OUP accepted manuscript

Lee

et al. 2016

Nucleic Acids Research

View full text Add to dashboard Cite

show abstract

“…TRAP uses method that combines the ORA and PT-based approaches to find significant pathways from KEGG pathway database and it is also designed for time-series gene expression data. For the literature search, Biomedical Entity Search Tool (BEST) [26] is used. BEST uses the concept of Maximal Coherent Semantic Unit for indexing keywords to associate the keyword and literatures efficiently.…”

Section: Methodsmentioning

confidence: 99%

Prioritizing biological pathways by recognizing context in time-series gene expression data

Lee

et al. 2016

BMC Bioinformatics

Self Cite

View full text Add to dashboard Cite

BackgroundThe primary goal of pathway analysis using transcriptome data is to find significantly perturbed pathways. However, pathway analysis is not always successful in identifying pathways that are truly relevant to the context under study. A major reason for this difficulty is that a single gene is involved in multiple pathways. In the KEGG pathway database, there are 146 genes, each of which is involved in more than 20 pathways. Thus activation of even a single gene will result in activation of many pathways. This complex relationship often makes the pathway analysis very difficult. While we need much more powerful pathway analysis methods, a readily available alternative way is to incorporate the literature information.ResultsIn this study, we propose a novel approach for prioritizing pathways by combining results from both pathway analysis tools and literature information. The basic idea is as follows. Whenever there are enough articles that provide evidence on which pathways are relevant to the context, we can be assured that the pathways are indeed related to the context, which is termed as relevance in this paper. However, if there are few or no articles reported, then we should rely on the results from the pathway analysis tools, which is termed as significance in this paper. We realized this concept as an algorithm by introducing Context Score and Impact Score and then combining the two into a single score. Our method ranked truly relevant pathways significantly higher than existing pathway analysis tools in experiments with two data sets.ConclusionsOur novel framework was implemented as ContextTRAP by utilizing two existing tools, TRAP and BEST. ContextTRAP will be a useful tool for the pathway based analysis of gene expression data since the user can specify the context of the biological experiment in a set of keywords. The web version of ContextTRAP is available at http://biohealth.snu.ac.kr/software/contextTRAP.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-1335-8) contains supplementary material, which is available to authorized users.

show abstract

“…BEST is a biomedical entity search tool for knowledge discovery from biomedical literature (Lee et al, 2016). Although PubMed (the free public interface to MEDLINE, which provides access to bibliographic information in MEDLINE as well as additional life science journals) provides a starting point to researchers, it only provides lists of relevant articles, leaving the task of extracting required information to the researchers themselves.…”

Section: Related Workmentioning

confidence: 99%

Entity-Centric Information Access with the Human-in-the-Loop for the Biomedical Domains

Yimam¹,

Remus²,

Panchenko³

et al. 2017

Proceedings of the Biomedical NLP Workshop

View full text Add to dashboard Cite

In this paper, we describe the concept of entity-centric information access for the biomedical domain. With entity recognition technologies approaching acceptable levels of accuracy, we put forward a paradigm of document browsing and searching where the entities of the domain and their relations are explicitly modeled to provide users the possibility of collecting exhaustive information on relations of interest. We describe three working prototypes along these lines: NEW/S/LEAK, which was developed for investigative journalists who need a quick overview of large leaked document collections; STORYFINDER, which is a personalized organizer for information found in web pages that allows adding entities as well as relations, and is capable of personalized information management; and adaptive annotation capabilities of WEBANNO, which is a general-purpose linguistic annotation tool. We will discuss future steps towards the adaptation of these tools to biomedical data, which is subject to a recently started project on biomedical knowledge acquisition. A key difference to other approaches is the centering around the user in a Human-in-theLoop machine learning approach, where users define and extend categories and enable the system to improve via feedback and interaction.

show abstract

BEST: Next-Generation Biomedical Entity Search Tool for Knowledge Discovery from Biomedical Literature

Cited by 90 publications

References 22 publications

OUP accepted manuscript

OUP accepted manuscript

Prioritizing biological pathways by recognizing context in time-series gene expression data

Entity-Centric Information Access with the Human-in-the-Loop for the Biomedical Domains

Contact Info

Product

Resources

About