Recognition of Disease Genetic Information from Unstructured Text Data Based on BiLSTM-CRF for Molecular Mechanisms

Disease relevant entities are an important task in mining unstructured text data from the biomedical literature for achieving knowledge. Autism spectrum disorder (ASD) is a disease related to neurological and developmental characterized by deficits communication social interaction repetitive behaviour. However, this kind of remains unclear date. In study, it identifies associated with using machine learning computational way collection molecular mechanisms ASD. Entities extracted autism deep bidirectional long short-term memory (BiLSTM) conditional random field (CRF) model. Compared other previous works, approach promising identifying disease. The proposed including five types evaluated GENIA corpus obtain F-score 76.81%. work has 9146 proteins, 145 RNAs, 7680 DNAs, 1058 cell-types, 981 cell-lines after removing repeated entities. Finally, we perform GO KEGG analyses test dataset. This study could serve as reference further studies on etiology basis provide explore genetic information.