Fig. 6
From: Optimized CRISPR guide RNA design for two high-fidelity Cas9 variants by deep learning
DeepHF model for gRNA activity prediction. The original gRNA sequence is first encoded and then embedded to get a new representation. This new representation is further processed by a BiLSTM to get the final representation which is then concatenated with hand-crafted features to serve as the input of the fully connected layer to be nonlinearly transformed. Finally, a linear transformation is performed to get the prediction score
