DeepARC:An Attention based Hybrid Model for Predicting Transcription Factor Binding Site from Positional Embedded DNA Sequence
1.data:the demo dataset
2.heat:visualizing the attention weight for sequence
3.load_data.py:preprocessing the data
4.models.py:deeparc model
5.myDataSet.py:including data preprocess and embedding
6.util.py: utils from sun
7.main.py:train and test the model
- download the training data to path
data/rawdata
(demo data has been provided in Repository) - run
main.py
to start training
DeepARC can be visited at http://www.deeparc.denglab.org/