Article
Biochemical Research Methods
Haoyu Zhang, Quan Zou, Ying Ju, Chenggang Song, Dong Chen
Summary: This study presents a novel model based on sequence distance matrix and support vector machine (SVM) for predicting DNA 6mA modification. The model achieved high accuracy rates and correlation coefficients on rice and mouse data, showing significant advantages over traditional machine learning methods.
CURRENT BIOINFORMATICS
(2022)
Review
Biotechnology & Applied Microbiology
Md Mehedi Hasan, Watshara Shoombuatong, Hiroyuki Kurata, Balachandran Manavalan
Summary: DNA N6-methyladenosine (6mA) methylation is an important epigenetic modification that plays key roles in biological processes, and accurate prediction of 6mA sites using bioinformatics tools can vary in performance across different species.
BRIEFINGS IN FUNCTIONAL GENOMICS
(2021)
Review
Biochemical Research Methods
Ke Han, Jianchun Wang, Yu Wang, Lei Zhang, Mengyao Yu, Fang Xie, Dequan Zheng, Yaoqun Xu, Yijie Ding, Jie Wan
Summary: This article introduces methods for predicting DNA N6-methyladenine sites and categorizes and analyzes traditional machine learning and deep learning methods. The authors reviewed existing model architectures, summarized and compared the results, providing guidance for subsequent researchers to choose suitable methods.
BRIEFINGS IN BIOINFORMATICS
(2023)
Review
Genetics & Heredity
Hao Li, Ning Zhang, Yuechen Wang, Siyuan Xia, Yating Zhu, Chen Xing, Xuefeng Tian, Yinan Du
Summary: DNA methylation is an important epigenetic mark in various biological activities. While past research focused on 5 mC, there has been limited attention given to N6-methyladenine (6 mA). With the development of detection technologies, 6 mA has been found in several eukaryotes. However, due to its low density and detection limitations, the prevalence and role of 6 mA in eukaryotic organisms are still debated.
FRONTIERS IN GENETICS
(2022)
Article
Biochemical Research Methods
Nguyen Quoc Khanh Le, Quang-Thai Ho
Summary: This study presents a novel model based on the transformer architecture and deep learning to accurately identify DNA 6mA sites from cross-species genomes. The model achieved excellent performance in independent testing, with an accuracy of 79.3% and a Matthews correlation coefficient (MCC) of 0.58. It outperformed baseline models and existing predictors, demonstrating the effectiveness of the proposed hybrid framework.
Review
Cell Biology
Xuwen Li, Zijian Zhang, Xinlong Luo, Jacob Schrier, Andrew D. Yang, Tao P. Wu
Summary: N-6-methyladenine (N-6-mA, m(6)dA, or 6mA) is a prevalent DNA modification in prokaryotes and has been recently identified in higher eukaryotes, such as mammals. It can serve as an epigenetic mark and play critical roles in various biological processes, but the function and regulatory mechanism of 6mA in eukaryotes, especially mammals, are still poorly understood and require further research.
Article
Biochemical Research Methods
Jianhua Cai, Guobao Xiao, Ran Su
Summary: In this study, a neural network-based bioinformatics model, GC6mA-Pred, is proposed to predict N6-methyladenine modifications in DNA sequences. The model extracts information from both sequence and graph levels and shows better performance on a newly built dataset.
Article
Biochemistry & Molecular Biology
Brian M. Debo, Benjamin J. Mallory, Andrew B. Stergachis
Summary: Low-level DNA N-6-methyldeoxyadenosine (DNA-m6A) has been recently discovered in various eukaryotes. However, the commonly used anti-m6A antibody-based methods for measuring DNA-m6A levels are known to be affected by DNA secondary structures, RNA contamination, and bacterial contamination. This study introduces an approach to validate the selectivity of antibody-based DNA-m6A methods and raises caution about their use for endogenous m6A quantification and mapping in eukaryotes.
Review
Chemistry, Multidisciplinary
Yuwei Sheng, Meijuan Zhou, Changjun You, Xiaoxia Dai
Summary: DNA methylation is a key type of DNA modification that plays important roles in biological processes. Recent studies have identified N-6-methyladenine (6mA) as an internal DNA modification that occurs dynamically in various eukaryotes, including humans. Increasing evidence suggests that 6mA may act as a novel epigenetic modification involved in the regulation of development, stress response, and diseases such as cancer and neurodegenerative disorders. This review summarizes the recent advances in detecting and studying the functional effects of 6mA modification, focusing on its biological consequences, relevance to human health, and dynamic regulation by methyltransferases, demethylases, and 6mA-binding proteins. Further research on the chemical and biological aspects of 6mA modification is expected to provide a better understanding of its potentially important roles in normal and pathological biological processes.
CHINESE CHEMICAL LETTERS
(2022)
Article
Biochemical Research Methods
Mengya Liu, Zhan-Li Sun, Zhigang Zeng, Kin-Man Lam
Summary: This paper proposes a novel deep learning method, MGF6mARice, for predicting 6mA sites in rice by devising DNA molecular graph feature and residual block structure. Experimental results show that this method outperforms existing approaches in 6mA prediction.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biochemistry & Molecular Biology
Sho Tsukiyama, Md Mehedi Hasan, Hiroyuki Kurata
Summary: N6-methyladenine (6mA) plays crucial roles in various epigenetic processes and diseases. To understand these mechanisms, researchers have developed a CNN-based 6mA site predictor, CNN6mA, which outperforms existing models and provides intelligible interpretation of the prediction mechanism through new architectures.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2023)
Article
Genetics & Heredity
Sara B. Fernandes, Nathalie Grova, Sarah Roth, Radu Corneliu Duca, Lode Godderis, Pauline Guebels, Sophie B. Meriaux, Andrew I. Lumley, Pascaline Bouillaud-Kremarik, Isabelle Ernens, Yvan Devaux, Henri Schroeder, Jonathan D. Turner
Summary: DNA methylation is a crucial epigenetic modification linked to gene regulation and non-malignant diseases. The presence of 6-methyladenine (6mA) in eukaryotes suggests its significance particularly during sensitive periods like embryogenesis. Studies in zebrafish and mice show an increase in 6mA levels during early development, with potential environmental influences leading to changes in this epigenetic mark.
FRONTIERS IN GENETICS
(2021)
Article
Plant Sciences
Zhixia Teng, Zhengnan Zhao, Yanjuan Li, Zhen Tian, Maozu Guo, Qianzi Lu, Guohua Wang
Summary: This article presents a novel model named i6mA-vote for predicting 6mA sites in plants. The model utilizes different coding strategies for feature extraction and combines multiple base-classifiers under a majority voting strategy. The performance of i6mA-vote is evaluated on three different plant genomes, showing its effectiveness in predicting 6mA sites across species. The results also highlight the importance of nucleotide position in identifying 6mA sites.
FRONTIERS IN PLANT SCIENCE
(2022)
Review
Biotechnology & Applied Microbiology
Hao Lv, Fu-Ying Dao, Dan Zhang, Hui Yang, Hao Lin
Summary: DNA modification is crucial in regulating gene expression in cell development, with advances in DNA sequencing technology allowing for the resolution of different modifications at a genome-wide scale. This has led to the discovery of new insights into the complexity and functions of multiple methylations. The review discusses various mapping approaches and the development of future sequencing technologies for improving detection resolution.
BIOTECHNOLOGY AND BIOENGINEERING
(2021)
Article
Cell Biology
Chengchuan Ma, Tingling Xue, Qi Peng, Jie Zhang, Jialiang Guan, Wanqiu Ding, Yi Li, Peixue Xia, Liankui Zhou, Tianyu Zhao, Sheng Wang, Li Quan, Chuan-Yun Li, Ying Liu
Summary: N-6-Methyldeoxyadenine (6mA) is a DNA modification found in metazoans that has potential biological function. Researchers have discovered that the levels of genomic 6mA change in response to pathogenic infection in Caenorhabditis elegans. The methyltransferase METL-9 has been identified as the enzyme responsible for catalyzing DNA 6mA modifications during pathogen infection. Deficiency of METL-9 impairs the induction of innate immune response genes and increases susceptibility to pathogen infection. These findings demonstrate that 6mA is a functional DNA modification involved in immunomodulation in C. elegans.
Article
Biochemical Research Methods
Jhabindra Khanal, Hilal Tayara, Quan Zou, Kil To Chong
Summary: In this study, a new deep learning model, DeepCap-Kcr, was proposed for accurate prediction of Kcr sites in proteins. The model outperformed existing methods and could learn internal hierarchical representations and important features from a small number of samples.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Chemistry, Multidisciplinary
Haider Kamran, Muhammad Tahir, Hilal Tayara, Kil To Chong
Summary: Enhancers are short motifs with high position variability and play an important role in gene regulation. Identification of enhancers is challenging due to their complexity, but recent advancements in computational tools and deep learning frameworks have shown comparable results with state-of-the-art methodologies.
APPLIED SCIENCES-BASEL
(2022)
Article
Biochemistry & Molecular Biology
Keerthana Jaganathan, Mobeen Ur Rehman, Hilal Tayara, Kil To Chong
Summary: In this study, an explainable machine-learning model was proposed to classify compounds with mitochondrial toxicity and non-toxicity. After experiments, the model achieved high prediction accuracy and showed significant improvement compared to existing methods.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2022)
Review
Biochemistry & Molecular Biology
Thi Tuyet Van Tran, Hilal Tayara, Kil To Chong
Summary: Drug distribution is a crucial process in pharmacokinetics, as it affects the effectiveness and safety of the drug. Lack of efficacy and uncontrollable toxicity are the major causes of drug failures. Advances in drug distribution property prediction, particularly through in silico methods, have reduced screening time and costs. This study provides comprehensive knowledge on drug distribution, including influencing factors and artificial intelligence-based prediction models. The review also presents future challenges and research directions, aiming to facilitate innovative approaches in drug discovery.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2023)
Article
Chemistry, Multidisciplinary
Waciar Ahmad, Hilal Tayara, Kil To Chong
Summary: Drug discovery (DD) research aims to discover new medications. Solubility is an important property in drug development. Aqueous solubility (AS) is a key attribute required for API characterization. In this study, deep learning models were created to predict the solubility of a wide range of molecules using the largest currently available solubility data set. The models were trained and tested on 9943 compounds, with the AttentiveFP-based network model outperforming on 62 anticancer compounds.
Article
Biochemical Research Methods
Prem Singh Bist, Hilal Tayara, Kil To Chong
Summary: We developed a computational model that accurately identifies viral escape mutational sequences based on natural language processing and prior knowledge of experimentally validated escape mutants. This model can be applied to other viruses using knowledge of escape mutants and protein sequence datasets.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Mobeen Ur Rehman, Hilal Tayara, Kil To Chong
Summary: In this study, a novel tool called DL-m6A is proposed for the identification of m6A sites in mammals using deep learning based on different encoding schemes. The tool utilizes three encoding schemes to provide contextual feature representation to the input RNA sequence. The results demonstrate that the proposed tool outperforms existing tools and can be of great use for biology experts.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Review
Chemistry, Medicinal
Thi Tuyet Van Tran, Agung Surya Wibowo, Hilal Tayara, Kil To Chong
Summary: Toxicity prediction in drug discovery is crucial for identifying safe and effective compounds, reducing late-stage failures. Artificial intelligence has shown promise in improving drug toxicity prediction through accurate and efficient methods. This review provides an overview of recent advances in AI-based drug toxicity prediction and highlights challenges and future perspectives, aiding researchers in understanding toxicity prediction and advancing drug discovery methods.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2023)
Review
Pharmacology & Pharmacy
Thi Tuyet Van Tran, Hilal Tayara, Kil To Chong
Summary: Drug metabolism and excretion are crucial in determining drug efficacy and safety. Artificial intelligence (AI) has emerged as a powerful tool for predicting these processes, offering potential for faster drug development and improved success rates. This review highlights recent advancements in AI-based prediction of drug metabolism and excretion, including deep learning and machine learning algorithms. It also provides a list of public data sources and prediction tools, discusses challenges in AI model development, and explores future perspectives in the field.
Article
Biochemical Research Methods
Zeeshan Abbas, Mobeen Ur Rehman, Hilal Tayara, Kil To Chong
Summary: In this article, a unique artificial intelligence-based technique called ORI-Explorer is developed to recognize origins of replication sites (ORIs) in four different eukaryotic species. ORI-Explorer combines multiple feature engineering techniques and utilizes the CatBoost Classifier. It outperforms existing predictors and provides key insights into model success through the SHapley Additive exPlanation method. ORI-Explorer aims to aid community-wide efforts in discovering potential ORIs and developing verifiable biological hypotheses.
Article
Biology
Priyash Dhakal, Hilal Tayara, Kil To Chong
Summary: We developed a stacking classifier algorithm that surpasses previous algorithms in predicting functional miRNA targets by effectively selecting conservative candidate target sites using feature encoding techniques.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Review
Chemistry, Medicinal
Thi Tuyet Van Tran, Hilal Tayara, Kil To Chong
Summary: Drug absorption is a crucial aspect in pharmaceutical research and development, and its prediction using in silico methods, particularly artificial intelligence, has shown promising results in reducing time and cost for screening drug candidates. This report provides an overview of recent studies on predicting absorption properties and highlights challenges and future directions in this field.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2023)
Article
Biochemical Research Methods
Sehi Park, Mobeen Ur Rehman, Farman Ullah, Hilal Tayara, Kil To Chong, Inanc Birol
Summary: The study developed positional features for predicting CpG site methylation patterns, using optimized classifiers and ensemble learning approaches. The CatBoost algorithm followed by the stacking algorithm outperformed existing DNA methylation identifiers. The proposed iCpG-Pos approach offers both accuracy and efficiency, making it a promising tool for advancing DNA methylation research and its applications in human health and well-being.
Article
Biochemistry & Molecular Biology
Syed Danish Ali, Hilal Tayara, Kil To Chong
Summary: piRNAs play a crucial role in maintaining genome integrity, and piRDA is an effective deep learning method for identifying piRNA-disease associations, facilitating drug development.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2022)
Article
Biochemical Research Methods
Syed Danish Ali, Waleed Alam, Hilal Tayara, Kil To Chong
Summary: piRNAs are a class of small RNAs that play important roles in maintaining germline cells, gene stability, and genome integrity, and are associated with various cancers. A predictor based on a deep learning architecture has been proposed, showing significant improvements in piRNA prediction and target mRNA deadenylation compared to existing computational methods.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2022)
Article
Biotechnology & Applied Microbiology
Zutan Li, Bingbing Jin, Jingya Fang
Summary: In this study, we propose MetaAc4C, an advanced deep learning model for accurate identification of N4-acetylcytidine (ac4C) sites using pre-trained BERT and various optimization techniques. By adapting generative adversarial networks to address data imbalance and augmenting training RNA samples, our model outperforms existing methods in terms of ACC, MCC, and AUROC.