Constructing a molecular interaction network for thyroid cancer via large-scale text mining of gene and pathway events

Abstract

BackgroundBiomedical studies need assistance from automated tools and easily accessible data to address the problem of the rapidly accumulating literature. Text-mining tools and curated databases have been developed to address such needs and they can be applied to improve the understanding of molecular pathogenesis of complex diseases like thyroid cancer.ResultsWe have developed a system, PWTEES, which extracts pathway interactions from the literature utilizing an existing event extraction tool (TEES) and pathway named entity recognition (PathNER). We then applied the system on a thyroid cancer corpus and systematically extracted molecular interactions involving either genes or pathways. With the extracted information, we constructed a molecular interaction network taking genes and pathways as nodes. Using curated pathway information and network topological analyses, we highlight key genes and pathways involved in thyroid carcinogenesis.ConclusionsMining events involving genes and pathways from the literature and integrating curated pathway knowledge can help improve the understanding of molecular interactions of complex diseases. The system developed for this study can be applied in studies other than thyroid cancer. The source code is freely available online at https://github.com/chengkun-wu/PWTEES.

Development of an information retrieval tool for biomedical patents

“…Due to the huge number of patents generated every year, there are several databases able to store these data. For instance, the World Intellectual Property Organization (WIPO) database had 2.7 million patents registered only in 2014 [3,21,22] .…”

“…Biomedical Text Mining (BioTM) is a sub-field of Text Mining that has been applied to biological and biomedical literature, seeking to automate the processing of the huge amounts of data that are generated every day including papers, reports and patents [1,2] , among others. Since these texts are written in natural language, in an unstructured form (without annotations about the text structure and available entities), searching for structured information by hand is quite time-consuming [3] . To automate this process, BioTM approaches become crucial, allowing the extraction of meaningful knowledge from texts, as a way to formulate scientific hypotheses more easily [4,5] .…”

Automated Extraction of Diagnostic Criteria From Electronic Health Records for Autism Spectrum Disorders: Development, Evaluation, and Application (Preprint)

“…The entities themselves have been predominantly single terms or relationships composed of single terms. For example, several projects focus, on annotating diseases or genes/proteins [13][14][15][16] and biomedical relationships between them [17,18] . When working with free text from EHR, a variety of entities have been the focus.…”

Development of Text Mining Tools for Information Retrieval from Patents

“…For instance, the WIPO database has 2.7 million patents registered only in 2014 [7][8][9]. Since these large amounts of data are available in an unstructured nature without annotations about the text structure and available entities, the search and extraction of relevant information is a difficult and time-consuming task, impossible to be done manually [7]. To exploit these data, automating that process, the Biomedical Text Mining (BioTM) field emerged [10].…”