Designs hand-crafted features without manual feature engineering. deepMiRGene is able to rediscover intrinsic features in a data-driven fashion. It is based on recurrent neural networks (RNNs) and specifically long short-term memory (LSTM) networks. This tool learns palindromic structures by using structure pre-processing of split and flip. It shows that it is possible to visually inspect the transition of the LSTM cell states on each position in the sequence in order to find existing structural features.