Predicting bipolar disorder and schizophrenia based on non-overlapping genetic phenotypes using deep neural network

摘要: Computational Psychiatry is an emerging field of science. It focuses on identifying the complex relationship between brain’s neurobiology. Mental illness has recently become important problem to be addressed as number people affected increasing over time. Schizophrenia and Bipolar Disorder are two major types psychiatric disorders. Most experienced these in their lifetime. But, diagnosing disorders even more a problem. Genetic factors play vital role developing mental illness. Interestingly, few have common genetic overlapping each other. causes detrimental effect accurately. To overcome this existing issue, Rank based Gene Biomarker Identification Classification framework proposed identify non-overlapping gene patterns bipolar disorder schizophrenia. The dataset used experiment obtained from Expression Omnibus database. As outcome experiment, seven biomarkers identified genes. Also, 60 68 informative schizophrenia feature subsets discriminate samples. Overlapping genes eliminated increase diagnostic accuracy performance system evaluated with standard machine learning algorithms. This attained 97.01% 95.65% Deep Neural Network model outperformed other benchmarked algorithms proved its efficacy.