Tarkhadkar, Sagar Santosh
MetadataShow full item record
Manual curation of knowledge from biomedical literature is both expensive and time consuming. Scientific publications in biomedicine have an enormous amount of valuable information on gene mutations and their impacts, which is significant in addressing multiple research problems. In this thesis, we have developed a text mining system for extracting and curating mutation impacts from full text scientific documents. The objective of this system is to populate biomedical knowledge-bases with accurate knowledge regarding mutation impacts, in a semi-automated way. We have used a number of Natural Language Processing tasks in developing this system. Furthermore, a curation module allows the scientists to decide if the mutation impact information is suitable to be included to the knowledge base, hence eliminating the possibility of adding incorrect data. Our prototype system has been used in the Protein Kinase domain, but can be adapted to work in other domains, in the future.