MPRA-DragoNN: Deciphering regulatory DNA sequences and noncoding genetic variants using neural network models of massively parallel reporter assays

This archive contains processed data and a trained model for MPRA-DragoNN, accompanying the paper “Deciphering regulatory DNA sequences and noncoding genetic variants using neural network models of massively parallel reporter assays” (Movva et al., PLoS ONE 2019). The repository includes pre-split HDF5 datasets (train/validation/test) derived from the Sharpr-MPRA dataset (Ernst et al., 2016), a Keras model architecture definition (model.json), and pretrained model weights (pretrained.hdf5). The model is a convolutional neural network trained to predict regulatory activity from 145 bp DNA sequences across multiple cellular contexts.