Fig. 5From: Prediction of RNA-protein sequence and structure binding preferences using deep convolutional and recurrent neural networksThe difference of predictive performance using CNN + BLSTM and only CNN. On the y-axis the performance of the full model with CNNs and BLSTM is shown. The x-axis shows the performance of the model using only the CNNs without BLSTM. The two red lines indicate the 2 times standard deviation of the difference between only using CNN and using CNN + BLSTMBack to article page