AUTHOR=Zhang Hanyu , Hung Che-Lun , Liu Meiyuan , Hu Xiaoye , Lin Yi-Yang 

TITLE=NCNet: Deep Learning Network Models for Predicting Function of Non-coding DNA

JOURNAL=Frontiers in Genetics

VOLUME=Volume 10 - 2019

YEAR=2019

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2019.00432

DOI=10.3389/fgene.2019.00432

ISSN=1664-8021

ABSTRACT=In human genome, it consists of 98.5% non-coding DNA sequences, and most of them have no
known function. However, majority of disease-associated variants lie in these regions. To predict
function of non-coding DNA is critical. Hence, we propose NCNet, which integrates deep residual
learning and sequence to sequence learning network, to predict transcription factor (TF) binding
sites, which can then be used to predict non-coding function. In NCNet, deep residual learning
is used to identify regulatory patterns of motifs and sequence to sequence learning is used to
extract sequential dependency between these patterns. With the identity shortcut technique,
NCNet achieves significant improvement compared to previews hybrid framework models for
identifying regulatory markers.