PubTator: a web-based text mining tool for assisting biocuration
By Chih-Hsuan Wei, Hung-Yu kao, Zhiyong Lu. 2013
586 citations recorded by the source corpus.
Manually curating knowledge from biomedical literature into structured databases is highly expensive and time-consuming, making it difficult to keep pace with the rapid growth of the literature. There is therefore a pressing need to assist biocuration with automated text mining tools. Here, we describe PubTator, a web-based system for assisting biocuration. PubTator is different from the few existing tools by featuring a PubMed-like interface, which many biocurators find familiar, and being equipped with multiple challenge-winning text mining algorithms to ensure the quality of its automatic results. Through a formal evaluation with two external user groups, PubTator was shown to be capable of improving both the efficiency and accuracy of manual curation. PubTator is publicly available at http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/Demo/PubTator/.
Subfields: Artificial Intelligence, Genetics, Molecular Biology, Information Extraction, Computational Biology & Bioinformatics
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Mohit Bansal
- Christopher R'e
- Luke S. Zettlemoyer
- Jianfeng Gao
- Percy Liang
- Yejin Choi
- Quoc V. Le
- Graham Neubig
In Genetics
- Robert Leaman
- Zhiyong Lu
- Rezarta Islamaj Dogan
- Chih-Hsuan Wei
- C. von Mering
- Lars Juhl Jensen
- Damian Szklarczyk
- Joanna S. Amberger
In Molecular Biology
- Zhiyong Lu
- Robert Leaman
- Peer Bork
- Chih-Hsuan Wei
- Lars Juhl Jensen
- C. von Mering
- Christopher R'e
- Damian Szklarczyk
In Information Extraction
- Robert Leaman
- Rezarta Islamaj Dogan
- Zhiyong Lu
- Heng Ji
- Chih-Hsuan Wei
- Pat Verga
- Jurica Seva
- Andrew McCallum
In Computational Biology & Bioinformatics
- Robert Leaman
- Rezarta Islamaj Dogan
- Zhiyong Lu
- Chih-Hsuan Wei
- Alexander J. Ratner
- Sampo Pyysalo
- Christopher R'e
- Goran Nenadic
Citations in the corpus
- PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts (2022, background): “Since then, many Web-based systems for annotating text have been developed (Stenetorp et al., 2012; Salgado et al., 2012; Wei et al., 2013; Yimam et al., 2013; Chen and Styler, 2013; Eckart de Castilho et al., 2016; Putra et al., 2020).”
- PaperRobot: Incremental Draft Generation of Scientific Ideas (2019, methodology): “We apply an entity mention extraction and linking system (Wei et al., 2013) to extract mentions of three entity types ( Disease , Chemical and Gene ) which are the core data categories in the Comparative Toxicogenomics Database (CTD) (Davis et al., 2016), and obtain a Medical Subject Head-ings…”
- tmChem: a high performance approach for chemical named entity recognition and normalization (2015, methodology): “Moreover, the results of a full run of Model 2 over PubMed is available in PubTator [6,38], including normalization using the lexical approach described in the Methods section.”
- The STRING database in 2021: customizable protein–protein networks, and functional characterization of user-uploaded gene/measurement sets (2020, methodology): “Specifically, we have created a high-confidence dataset consisting of millions of protein and non-protein text spans, which three widely adopted protein name recognition systems (54–56) agreed on tagging as such.”
- OMIM.org: leveraging knowledge across phenotype–gene relationships (2018, background): “Review of high-impact journals, targeted searches of PubMed for unique phenotype– gene relationships, user-provided suggestions, and articles identified in the curation process are tracked and computationally tagged for gene, disease, and mutation concepts using PubTator (3).”
- Deep learning meets ontologies: experiments to anchor the cardiovascular disease ontology in the biomedical literature (2018, background): “There are, however, studies that have a list of multiple types of biomedical entities, such as PubTator [109], and BEST [110].”
- BioCreative V CDR task corpus: a resource for chemical disease relation extraction (2016, background): “Due to the high cost of manual curation and rapid growth of the biomedical literature, several attempts have been made to assist curation using text-mining systems (4,5) including the automatic extraction of CDRs (6).”
- DNorm: disease name normalization with pairwise learning to rank (2013, methodology): “We have applied DNorm to all PubMed abstracts and made the results publicly available in PubTator (Wei et al. , 2012, 2013).”
- Improving chemical disease relation extraction with rich features and weakly labeled data (2016, unknown): “Automatic text-mining tools have been proposed to assist the manual creation [34, 45, 54] and/or to directly generate large-scale results for computational purposes [47, 49].”
- Marginal Likelihood Training of BiLSTM-CRF for Biomedical Named Entity Recognition from Disjoint Label Sets (2018, methodology): “In these approaches, titles and abstracts from PubMed are annotated using Pubtator, a state of the art entity tagging and linking/normalization system (Wei et al., 2013).”
- PubTator central: automated concept annotation for biomedical full text articles (2019, background): “PubTator (10) was one of the first such systems to provide automated concept annotations of several important biomedical concept types – genes / proteins, genetic variants, diseases, chemicals, and species – across all PubMed article abstracts.”
- Data programming with DDLite: putting humans in a different part of the loop (2016, background): “DDLite provides a library of general candidate extraction operators.”
- An extensive review of tools for manual annotation of documents (2019, background): “Other nonschematic tools include MDSWriter (designed for the annotation of summaries), Textpresso (suitable for data curation) and PubTator (limited to a few biomedical semantic types).”
- A survey of current trends in computational drug repositioning (2016, background): “The biomedical and pharmaceutical knowledge available in literature [70, 71] or databases contains vast amount of information for drugs and diseases, which can be automatically mined [72–74] and retrieved [75], thanks to the recent advances in text mining research [76].”
- NCBI disease corpus: A resource for disease name recognition and concept normalization (2014, methodology): “Screenshot of our annotation software PubTator displaying the differences in annotation for two annotators at the end of Phase I. R.I. Dog ˘an et al. / Journal of Biomedical Informatics 47 (2014) 1–10 5 fact that the long form of the disease is usually defined elsewhere in the same abstract .”