4mCCNN: Identification of N4-Methylcytosine Sites in Prokaryotes Using Convolutional Neural Network

• Main Authors: Jhabindra Khanal, Iman Nazari, Hilal Tayara, Kil To Chong
• Format: Article
• Language: English
• Published: IEEE 2019-01-01
• Series: IEEE Access
• Subjects: Convolutional neural network; DNA methylation; DNA N4 – methylcytosine(4mC); sequence analysis
• Online Access: https://ieeexplore.ieee.org/document/8846679/

The epigenetic modification, DNA N4 - methylcytosine(4mC) plays an important role in DNA expression, repair, and replication. It simply plays a crucial role in restriction-modification systems. The better and accurate prediction of 4mC sites in DNA is much-needed work to understand their functional behavior that leads to help in both drug discovery and biomedical research. Therefore, an accurate computational model is required. In this work, we present an efficient one-dimensional convolutional neural network (CNN) model, called 4mCCNN, for 4mc sites identifications in Caenorhabditis elegans, Drosophila melanogaster, Arabidopsis thaliana, Escherichia coli, Geoalkalibacter subterraneus, and Geobacter pickeringii. Existing methods were developed by machine learning algorithms for identifying the 4mc sites using handcrafted features, while the proposed model extracts the features of the 4mC sites from DNA sequence automatically using the CNN model. The performance of the proposed model has been evaluated on benchmark datasets and achieved generally better outcomes in identifying 4mc sites as compared to the state-of-the-art predictors. The developed 4mCNN model was constructed in a web server at https://home.jbnu.ac.kr/NSCL/4mCCNN.htm.