Predict microRNA target by Convolutional Neural Network.
The input data should consist of the values of the complementary features, the accessibility features and the conservation related features of candidate target pairs.
The online tool is trained with sequences of human miRNAs and mRNAs. There are significant differences of miRNA target mechanism between different species.
So we strongly suggest users download the source code and train it on their own PC using their specific training data and parameters.
The positive/negative candidate target pairs, the corresponding feature values and the CNN code used in our work are freely available in "Downloads".