Accepted_test

Major depressive disorder affects the quality of life for millions of people. Various reasons  of this disorder were proposed, and the full range of risk factors are under study. The aim of this work is to study possible molecular mechanisms of major depressive disorder pathogenesis  and predisposing psychological factors based on text mining approach. A variety of machine learning methods (naїve Bayes, conditional random fields, artificial neural networks) as well as regular expressions and dictionaries for named entity recognition tasks were applied for an analysis of epigenetic mechanisms, gene and proteins involved in the development of the major depressive disorder. In particular, based on the text-mining approach we identitied factors included in estrogen and glucocorticoid receptor regulation as important for the development of major depression. An important result of the work is a list of environmental conditions, especially in childhood, which, according to the results of the analysis, may contribute to the development of major depressive disorder in adults.