PubTator is developed based on a prototype system that was previously used at the NCBI for various manual curation projects such as annotating disease mentions in PubMed abstracts. In response to call for participation in BioCreative 2012, we significantly extended our previous system in developing PubTator. First, relevance ranking and concept highlighting were added to ease the task of document triage. Second, state-of-the-art named entity recognition tools (e.g. winning gene normalization systems [2,3] in BioCreative III) were integrated to pre-tag bioconcepts of interest, as a way to facilitate the task of gene/disease/chemical annotation. Third, PubTator was developed to have a look-and-feel similar to PubMed, thus minimizing the learning efforts required for new users. Furthermore, a standard PubMed search option is made available in PubTator, which would allow our users to make a hassle-free move of their saved PubMed queries (a common practice for curators doing document triage) into this new curation system. Finally, by taking advantage of pre-tagging bioconcepts, PubTator also allows its users to do semantic search besides the traditional keyword based search, a novel feature not available in PubMed.
[1]
Xiaoyan Zhu,et al.
GeneTUKit: a software for document-level gene normalization
,
2011,
Bioinform..
[2]
Hung-Yu Kao,et al.
Cross-species gene normalization by species inference
,
2011,
BMC Bioinformatics.
[3]
Zhiyong Lu,et al.
Semi-automatic semantic annotation of PubMed queries: A study on quality, efficiency, satisfaction
,
2011,
J. Biomed. Informatics.
[4]
W. John Wilbur,et al.
Text Mining Techniques for Leveraging Positively Labeled Data
,
2011,
BioNLP@ACL.
[5]
Zhiyong Lu,et al.
SR4GN: A Species Recognition Software Tool for Gene Normalization
,
2012,
PloS one.