A TensorFlow implementation of DeepBind model is proposed. DeepBind is a deep learning model using convolutional neural networks used for predicting TF binding sites. The model was proposed in the paper "Predicting the sequence specificities of DNA and RNA-binding proteins by deep learning" published in 2015.
The architecture of the model and the calibration phase steps are explained in this figure from the paper:
