AUTHOR=Cao Linan , Liu Pei , Chen Jialong , Deng Lei 

TITLE=Prediction of Transcription Factor Binding Sites Using a Combined Deep Learning Approach

JOURNAL=Frontiers in Oncology

VOLUME=Volume 12 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2022.893520

DOI=10.3389/fonc.2022.893520

ISSN=2234-943X

ABSTRACT=In the process of regulating gene expression and evolution, such as DNA replication and mRNA transcription,
we found that the binding of transcription factors(TFs) to transcription factor binding sites(TFBS) plays a vital role.
Precisely modeling the specificity of gene and searching for TFBS is helpful to comprehend
the mechanism of cell expression. 
Over the years, computational and deep learning methods searching for TFBS have become an active field of research. However, these methods generally cannot meet high performance and interpretability at the same time.
Here, we develop an accurate and interpretable attention-based hybrid
approach, DeepARC that uses a convolutional neural network (CNN) and
recurrent neural network (RNN) to predict TFBS.
DeepARC employs a positional embedding method to extract the hidden embedding from a DNA sequence, including the positional information from OneHot encoding and the distributed embedding from DNA2Vec.
DeepARC feed the positional embedding of the DNA sequence into a CNN-BiLSTM-Attention-based framework to complete the task of finding the motif.
Take advantage of the attention mechanism, DeepARC is able to gain greater access to valuable information about the motif as well as bring interpretability to the work of searching for motif through the attention weight graph.
Moreover, DeepARC achieves promising performances with a 90.8\% average ROCAUC score on five cell lines(A549, GM12878, Hep-G2, H1-hESC, Hela) in ENCODE(Encyclopedia of DNA Elements) project.
We also compare the positional embedding with OneHot and DNA2Vec and gain a competitive advantage as a result.