Browsing by Author "Ajou University School of Medicine"

Filter results by typing the first few letters
Now showing 1 - 6 of 6
  • No Thumbnail Available
    PublicationMetadata only
    BERT4Bitter: A bidirectional encoder representations from transformers (BERT)-based model for improving the prediction of bitter peptides
    (2021-09-01) Phasit Charoenkwan; Chanin Nantasenamat; Md Mehedi Hasan; Balachandran Manavalan; Watshara Shoombuatong; Kyushu Institute of Technology; Ajou University School of Medicine; Tulane University School of Medicine; Mahidol University; Chiang Mai University
    Motivation: The identification of bitter peptides through experimental approaches is an expensive and timeconsuming endeavor. Due to the huge number of newly available peptide sequences in the post-genomic era, the development of automated computational models for the identification of novel bitter peptides is highly desirable. Results: In this work, we present BERT4Bitter, a bidirectional encoder representation from transformers (BERT)- based model for predicting bitter peptides directly from their amino acid sequence without using any structural information. To the best of our knowledge, this is the first time a BERT-based model has been employed to identify bitter peptides. Compared to widely used machine learning models, BERT4Bitter achieved the best performance with an accuracy of 0.861 and 0.922 for cross-validation and independent tests, respectively. Furthermore, extensive empirical benchmarking experiments on the independent dataset demonstrated that BERT4Bitter clearly outperformed the existing method with improvements of 8.0% accuracy and 16.0% Matthews coefficient correlation, highlighting the effectiveness and robustness of BERT4Bitter. We believe that the BERT4Bitter method proposed herein will be a useful tool for rapidly screening and identifying novel bitter peptides for drug development and nutritional research.
  • No Thumbnail Available
    PublicationMetadata only
    Critical evaluation of web-based DNA N6-methyladenine site prediction tools
    (2021-07-01) Md Mehedi Hasan; Watshara Shoombuatong; Hiroyuki Kurata; Balachandran Manavalan; Kyushu Institute of Technology; Ajou University School of Medicine; Mahidol University
    Methylation of DNA N6-methyladenosine (6mA) is a type of epigenetic modification that plays pivotal roles in various biological processes. The accurate genome-wide identification of 6mA is a challenging task that leads to understanding the biological functions. For the last 5 years, a number of bioinformatics approaches and tools for 6mA site prediction have been established, and some of them are easily accessible as web application. Nevertheless, the accurate genome-wide identification of 6mA is still one of the challenging works that lead to understanding the biological functions. Especially in practical applications, these tools have implemented diverse encoding schemes, machine learning algorithms and feature selection methods, whereas few systematic performance comparisons of 6mA site predictors have been reported. In this review, 11 publicly available 6mA predictors evaluated with seven different species-specific datasets (Arabidopsis thaliana, Tolypocladium, Diospyros lotus, Saccharomyces cerevisiae, Drosophila melanogaster, Caenorhabditis elegans and Escherichia coli). Of those, few species are close homologs, and the remaining datasets are distant sequences. Our independent, validation tests demonstrated that Meta-i6mA and MM-6mAPred models for A. thaliana, Tolypocladium, S. cerevisiae and D. melanogaster achieved excellent overall performance when compared with their counterparts. However, none of the existing methods were suitable for E. coli, C. elegans and D. lotus. A feasibility of the existing predictors is also discussed for the seven species. Our evaluation provides useful guidelines for the development of 6mA site predictors and helps biologists selecting suitable prediction tools.
  • No Thumbnail Available
    PublicationMetadata only
    Cultural validation and language translation of the scientific SCI exercise guidelines for use in Indonesia, Japan, Korea, and Thailand
    (2021-01-01) Yukio Mikami; Damayanti Tinduh; Kun Ho Lee; Chayaporn Chotiyarnwong; Jan W. van der Scheer; Kyung Su Jung; Hiroshi Shinohara; Inggar Narasinta; Seung Hyun Yoon; Napatpaphan Kanjanapanang; Takafumi Sakai; Martha K. Kusumawardhani; Jinho Park; Pannika Prachgosin; Futoshi Obata; Ditaruni Asrina Utami; Phairin Laohasinnarong; Indrayuni Lukitra Wardhani; Siraprapa Limprasert; Fumihiro Tajima; Victoria L. Goosey-Tolfrey; Kathleen A. Martin Ginis; Siriraj Hospital; Takarazuka University of Medical and Health Care; Department of Public Health and Primary Care; Universitas Airlangga; Ajou University School of Medicine; Wakayama Medical University; Aomori University of Health and Welfare; Texas A and M University System; University of British Columbia Okanagan; The University of British Columbia; Dankook University; Loughborough University
    Context: Indonesia, Japan, Korea, Thailand. Objective: To culturally validate and translate the Scientific Exercise Guidelines for Adults with Spinal Cord Injury (SEG-SCI) for use in four Asian countries. Design: Systematic Review Participants: N/A Methods: A systematic review was conducted to identify all published English- and local-language studies conducted in Indonesia, Japan, Korea, and Thailand, testing the effects of exercise training interventions on fitness and cardiometabolic health in adults with acute or chronic SCI. Protocols and results from high-quality controlled studies were compared with the SEG-SCI. Forward and backward translation processes were used to translate the guidelines into Bahasa Indonesian, Japanese, Korean and Thai languages. Results: Fifteen studies met the review criteria. At least one study from each country implemented exercise prescriptions that met or exceeded the SEG-SCI. Two were controlled studies. In those two studies, relative to control conditions, participants in exercise conditions achieved significant improvements in fitness or cardiometabolic health outcomes only when the exercise intervention protocol met or exceeded the SEG-SCI. During the language translation processes, end-users confirmed that SEG-SCI language and terminology were clear. Conclusion: Clinical researchers in Indonesia, Japan, Korea and Thailand have implemented exercise protocols that meet or exceed the SCI-SEG. Results of high-quality studies align with the SEG-SCI recommendations. Based on this evidence, we recommend that the SEG-SCI be adopted in these countries. The cultural validation and translation of the SEG-SCI is an important step towards establishing consistent SCI exercise prescriptions in research, clinical and community settings around the world.
  • No Thumbnail Available
    PublicationMetadata only
    NeuroPred-FRL: an interpretable prediction model for identifying neuropeptide using feature representation learning
    (2021-11-05) Md Mehedi Hasan; Md Ashad Alam; Watshara Shoombuatong; Hong Wen Deng; Balachandran Manavalan; Hiroyuki Kurata; Kyushu Institute of Technology; Ajou University School of Medicine; Japan Society for the Promotion of Science; Tulane University School of Medicine; Mahidol University
    Neuropeptides (NPs) are the most versatile neurotransmitters in the immune systems that regulate various central anxious hormones. An efficient and effective bioinformatics tool for rapid and accurate large-scale identification of NPs is critical in immunoinformatics, which is indispensable for basic research and drug development. Although a few NP prediction tools have been developed, it is mandatory to improve their NPs' prediction performances. In this study, we have developed a machine learning-based meta-predictor called NeuroPred-FRL by employing the feature representation learning approach. First, we generated 66 optimal baseline models by employing 11 different encodings, six different classifiers and a two-step feature selection approach. The predicted probability scores of NPs based on the 66 baseline models were combined to be deemed as the input feature vector. Second, in order to enhance the feature representation ability, we applied the two-step feature selection approach to optimize the 66-D probability feature vector and then inputted the optimal one into a random forest classifier for the final meta-model (NeuroPred-FRL) construction. Benchmarking experiments based on both cross-validation and independent tests indicate that the NeuroPred-FRL achieves a superior prediction performance of NPs compared with the other state-of-the-art predictors. We believe that the proposed NeuroPred-FRL can serve as a powerful tool for large-scale identification of NPs, facilitating the characterization of their functional mechanisms and expediting their applications in clinical therapy. Moreover, we interpreted some model mechanisms of NeuroPred-FRL by leveraging the robust SHapley Additive exPlanation algorithm.
  • No Thumbnail Available
    PublicationMetadata only
    StackIL6: a stacking ensemble model for improving the prediction of IL-6 inducing peptides
    (2021-11-05) Phasit Charoenkwan; Wararat Chiangjong; Chanin Nantasenamat; Md Mehedi Hasan; Balachandran Manavalan; Watshara Shoombuatong; Ramathibodi Hospital; Kyushu Institute of Technology; Ajou University School of Medicine; Mahidol University; Chiang Mai University
    The release of interleukin (IL)-6 is stimulated by antigenic peptides from pathogens as well as by immune cells for activating aggressive inflammation. IL-6 inducing peptides are derived from pathogens and can be used as diagnostic biomarkers for predicting various stages of disease severity as well as being used as IL-6 inhibitors for the suppression of aggressive multi-signaling immune responses. Thus, the accurate identification of IL-6 inducing peptides is of great importance for investigating their mechanism of action as well as for developing diagnostic and immunotherapeutic applications. This study proposes a novel stacking ensemble model (termed StackIL6) for accurately identifying IL-6 inducing peptides. More specifically, StackIL6 was constructed from twelve different feature descriptors derived from three major groups of features (composition-based features, composition-transition-distribution-based features and physicochemical properties-based features) and five popular machine learning algorithms (extremely randomized trees, logistic regression, multi-layer perceptron, support vector machine and random forest). To enhance the utility of baseline models, they were effectively and systematically integrated through a stacking strategy to build the final meta-based model. Extensive benchmarking experiments demonstrated that StackIL6 could achieve significantly better performance than the existing method (IL6PRED) and outperformed its constituent baseline models on both training and independent test datasets, which thereby support its excellent discrimination and generalization abilities. To facilitate easy access to the StackIL6 model, it was established as a freely available web server accessible at http://camt.pythonanywhere.com/StackIL6. It is anticipated that StackIL6 can help to facilitate rapid screening of promising IL-6 inducing peptides for the development of diagnostic and immunotherapeutic applications in the future.
  • No Thumbnail Available
    PublicationMetadata only
    Umpred-frl: A new approach for accurate prediction of umami peptides using feature representation learning
    (2021-12-01) Phasit Charoenkwan; Chanin Nantasenamat; Md Mehedi Hasan; Mohammad Ali Moni; Balachandran Manavalan; Watshara Shoombuatong; The University of Queensland; Ajou University School of Medicine; Tulane University School of Medicine; Mahidol University; Chiang Mai University
    Umami ingredients have been identified as important factors in food seasoning and production. Traditional experimental methods for characterizing peptides exhibiting umami sensory properties (umami peptides) are time-consuming, laborious, and costly. As a result, it is preferable to develop computational tools for the large-scale identification of available sequences in order to identify novel peptides with umami sensory properties. Although a computational tool has been developed for this purpose, its predictive performance is still insufficient. In this study, we use a feature representation learning approach to create a novel machine-learning meta-predictor called UMPred-FRL for improved umami peptide identification. We combined six well-known machine learning algorithms (extremely randomized trees, k-nearest neighbor, logistic regression, partial least squares, random forest, and support vector machine) with seven different feature encodings (amino acid composition, amphiphilic pseudo-amino acid composition, dipeptide composition, composition-transition-distribution, and pseudo-amino acid composition) to develop the final meta-predictor. Extensive experimental results demonstrated that UMPred-FRL was effective and achieved more accurate performance on the benchmark dataset compared to its baseline models, and consistently outperformed the existing method on the independent test dataset. Finally, to aid in the high-throughput identification of umami peptides, the UMPred-FRL web server was established and made freely available online. It is expected that UMPred-FRL will be a powerful tool for the cost-effective large-scale screening of candidate peptides with potential umami sensory properties.