MetadataShow full item record
The use of computing has become nearly ubiquitous in biomedical sciences, which resulted in the creation of numerous data sets and data sources. To assure accuracy, data sets such as UniProt, Reactome, and COSMIC have to be manually verified and curated by humans. Some data sets can be created by text mining or data extraction from text. However, due to the imprecise nature of these methods and the ambiguity of natural language, data produced by these techniques requires verification by humans. In this thesis, we describe a software system for the curation of mutation impacts extracted by text mining of published biomedical articles. Our curatorial system provides a significant level of assistance to the curators, implements a multi-level verification process and provides a feedback mechanism, which is used to improve the text mining system. The curated mutation impact data will be included in the ProKinO ontology, in the near future.