DeepCRISPR: optimized CRISPR guide RNA design by deep learning

DeepCRISPR is a deep learning based prediction model for sgRNA on-target knockout efficacy and genome-wide off-target cleavage profile prediction.

This model is based on a carefully designed hybrid deep neural network for model training and prediction.

Current version focuses on conventional NGG-based sgRNA design for SpCas9 in human species, for it is widely used in related experiments.