Systematic Analysis and Accurate Identification of DNA N4-Methylcytosine Sites by Deep Learning

Abstract

DNA N4-methylcytosine (4mC) is a pivotal epigenetic modification that plays an essential role in DNA replication, repair, expression and differentiation. To gain insight into the biological functions of 4mC, it is critical to identify their modification sites in the genomics. Recently, deep learning has become increasingly popular in recent years and frequently employed for the 4mC site identification. However, a systematic analysis of how to build predictive models using deep learning techniques is still lacking. In this work, we first summarized all existing deep learning-based predictors and systematically analyzed their models, features and datasets, etc. Then, using a typical standard dataset with three species (A. thaliana, C. elegans, and D. melanogaster), we assessed the contribution of different model architectures, encoding methods and the attention mechanism in establishing a deep learning-based model for the 4mC site prediction. After a series of optimizations, convolutional-recurrent neural network architecture using the one-hot encoding and attention mechanism achieved the best overall prediction performance. Extensive comparison experiments were conducted based on the same dataset. This work will be helpful for researchers who would like to build the 4mC prediction models using deep learning in the future.

A Grid Search-Based Multilayer Dynamic Ensemble System to Identify DNA N4—Methylcytosine Using Deep Learning Approach

“…In TNC, all samples of 41 nt produce 39 components with the equation of L − k + 1. Here, L stands for the sequence length, and k stands for the K-mer value as an integer [ 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 ]. ATG, TGC, GCG, and CGA are the four 3-mers that can be tokenized from the DNA sequence “ATGCGA,” for instance.…”

“…Yu. Lezheng et al (2022) [ 11 ] proposed a convolutional recurrent neural network model to identify DNA N4-methylcytosine. For representing DNA sequences, they considered one-hot and dictionary encoding methods.…”

Shaping eukaryotic epigenetic systems by horizontal gene transfer

“…Since 2017, over 30 publications aimed at characterizing 4mC distribution in eukaryotic genomes were devoted to continuous improvement of 4mC prediction accuracy using machine learning approaches ( [67,68] and references therein). However, the value of these methods is uncertain, as they were developed and trained on reference datasets [69,70] which have not been rigorously proven to harbor the 4mC mark.…”

Fast and Efficient Design of Deep Neural Networks for Predicting N7-Methylguanosine Sites Using autoBioSeqpy

“…Therefore, throughout the use of autoBioSeqpy, the users only need to provide the data sets and the model architecture codes. Recently, we have applied autoBioSeqpy to practice with good results, e.g., druggable protein prediction, 37 DL algorithm development for DNA N 4 methylcytosine, 38 anti-cancer peptide design, 39 and the distinction between bacterial type III and IV secreted effectors. 40 Hence, the present study will utilize autoBioSeqpy for the rapid and efficient creation, training, and utilization of DL models for the detection of m 7 G sites.…”