bsml320 / deepface

Deep learning model for predicting craniofacial variants

DeepFace

Deep learning model for predicting craniofacial variants ##DeepFace use basenji as the backbone model ##Install https://github.com/calico/basenji in virtual environment basenji

source activate basenji

##preprocess the bigWig file following these steps https://github.com/calico/basenji/blob/master/tutorials/preprocess.ipynb

##train model ! /PATH_to/DeepFace_train.py -o /PATH_to_save_model/ /PATH_to_params.json /PATH_to_processed_bigwig_data

##more details please check https://github.com/calico/basenji to do the prediction of your own data.

1787by204_SAD.csv contains the SAD score for 1787 OFC-related variants (input_1787_OFC_variants.txt) across 204 craniofacial development epigenomic profiles （OFC_204_profile_wigs.txt）.