I-DNAN6mA: Accurate Identification of DNA N⁶-Methyladenine Sites Using the Base-Pairing Map and Deep Learning

Abstract

The recent discovery of numerous DNA N6-methyladenine (6mA) sites has transformed our perception about the roles of 6mA in living organisms. However, our ability to understand them is hampered by our inability to identify 6mA sites rapidly and cost-efficiently by existing experimental methods. Developing a novel method to quickly and accurately identify 6mA sites is critical for speeding up the progress of its function detection and understanding. In this study, we propose a novel computational method, called I-DNAN6mA, to identify 6mA sites and complement experimental methods well, by leveraging the base-pairing rules and a well-designed three-stage deep learning model with pairwise inputs. The performance of our proposed method is benchmarked and evaluated on four species, i.e., Arabidopsis thaliana, Drosophila melanogaster, Rice, and Rosaceae. The experimental results demonstrate that I-DNAN6mA achieves area under the receiver operating characteristic curve values of 0.967, 0.963, 0.947, 0.976, and 0.990, accuracies of 91.5, 92.7, 88.2, 0.938, and 96.2%, and Mathew’s correlation coefficient values of 0.855, 0.831, 0.763, 0.877, and 0.924 on five benchmark data sets, respectively, and outperforms several existing state-of-the-art methods. To our knowledge, I-DNAN6mA is the first approach to identify 6mA sites using a novel image-like representation of DNA sequences and a deep learning model with pairwise inputs. I-DNAN6mA is expected to be useful for locating functional regions of DNA.

SDBA: Score Domain-Based Attention for DNA N4-Methylcytosine Site Prediction from Multiperspectives

“…Unlike the CNN-RNN serial processing in previous studies, , we used CNN and RNN to represent DNA sequences in parallel and then used SDBA to weight these representations. The advantages are that we could not only avoid the noise accumulation problem caused by the serial structure but also make the two models learn from the extracted features from the other models in the encoding process to achieve the purpose of joint learning.…”

Optimizing Hyperparameter Tuning in Machine Learning to Improve the Predictive Performance of Cross-Species N6-Methyladenosine Sites

“…6 Subsequent studies have led to the development of other genome 6 mA site identification tools such as MethyRNA, 12 M6AMRFS, 13 Le and Ho, 14 i6 mA-Caps, 15 This article is licensed under CC-BY 4 ENet-6 mA, 16 m6A-TSHub, 17 DeepM6ASeq-EL, 18 and I-DNAN6 mA. 19 The successful development of machine learning models for 6 mA prediction significantly facilitates our understanding of its functional roles and molecular mechanisms in various biological processes, thereby providing valuable insights into the development of novel therapeutic strategies. Inspired by previous successful research, we aim to further investigate various encoding methods and models for detecting 6 mA sites.…”

“…Moreover, Abbas et al applied one-hot encoding for genome data transformation and developed a deep neural network (namely, TS-m6A-DL) to classify 6 mA sites in humans, mice, and rats . Subsequent studies have led to the development of other genome 6 mA site identification tools such as MethyRNA, M6AMRFS, Le and Ho, i6 mA-Caps, ENet-6 mA, m6A-TSHub, DeepM6ASeq-EL, and I-DNAN6 mA …”

“… 6 Subsequent studies have led to the development of other genome 6 mA site identification tools such as MethyRNA, 12 M6AMRFS, 13 Le and Ho, 14 i6 mA-Caps, 15 ENet-6 mA, 16 m6A-TSHub, 17 DeepM6ASeq-EL, 18 and I-DNAN6 mA. 19 …”