![\"a72b5b66-69dd-4e7b-9ec1-aa300fabcb04\"](\"https:\/\/medicine.nus.edu.sg\/dbmi\/wp-content\/uploads\/sites\/27\/2026\/03\/a72b5b66-69dd-4e7b-9ec1-aa300fabcb04.png\" "\\"a72b5b66-69dd-4e7b-9ec1-aa300fabcb04\\"")
\n\nUnderstanding disease mechanisms at the molecular level is essential for advancing precision medicine. Research in this area develops computational approaches to analyse biological sequences, protein structures, and molecular interactions using modern deep learning architectures.<\/p>\n
Recent work has focused on\u00a0multimodal graph neural networks and transformer-based models for protein function prediction<\/strong>, integrating structural information derived from AlphaFold with sequence-based representations to improve biological inference. These approaches have been applied to problems such as\u00a0protein function annotation, peptide therapeutic discovery, and DNA methylation prediction<\/strong>, contributing to advances in computational biology and drug discovery.<\/p>\n Representative contributions include studies on\u00a0protein function prediction using graph-based deep learning models,\u00a0ensemble deep learning approaches for anticancer peptide prediction<\/strong>, and\u00a0deep neural network frameworks for identifying post-translational modification sites in proteins<\/strong>. These efforts demonstrate how computational intelligence can accelerate biological discovery and expand our understanding of complex molecular systems.<\/p>\n\t\t\t\t\t\t Understanding disease mechanisms at the molecular level is essential for advancing precision medicine. Research in this area develops computational approaches to analyse biological sequences, protein structures, and molecular interactions using modern deep learning architectures.<\/p>\n Recent work has focused on\u00a0multimodal graph neural networks and transformer-based models for protein function prediction<\/strong>, integrating structural information derived from AlphaFold with sequence-based representations to improve biological inference. These approaches have been applied to problems such as\u00a0protein function annotation, peptide therapeutic discovery, and DNA methylation prediction<\/strong>, contributing to advances in computational biology and drug discovery.<\/p>\n Representative contributions include studies on\u00a0protein function prediction using graph-based deep learning models,\u00a0ensemble deep learning approaches for anticancer peptide prediction<\/strong>, and\u00a0deep neural network frameworks for identifying post-translational modification sites in proteins<\/strong>. These efforts demonstrate how computational intelligence can accelerate biological discovery and expand our understanding of complex molecular systems.<\/p>\n\t\t\t\t\t\tMultimodal Clinical Intelligence<\/h4>\n\t\t\t
Digital Therapeutics and Behavioural Health<\/h4>\n\t\t\tBeyond diagnosis and prediction, an important focus of Computational Health Intelligence is the development of\u00a0digital interventions that actively improve health outcomes<\/b>. This includes intelligent systems that combine artificial intelligence with mobile health technologies, interactive applications, and sensor-based monitoring to support rehabilitation and lifestyle medicine.\n\u00a0\nA major research direction involves the design and evaluation of\u00a0AI-powered exergames and digital rehabilitation platforms<\/b>\u00a0aimed at promoting healthy ageing. For example, the\u00a0SinDance<\/i>\u00a0project developed a culturally contextualised dance-based exergame designed to enhance physical function and well-being among older adults. This work has been evaluated through clinical studies and randomised trials demonstrating improvements in physical and psychosocial outcomes.\n\u00a0\nComplementary work has also explored\u00a0mobile health interventions for cardiac rehabilitation, stroke recovery, and weight management<\/b>, including smartphone-based platforms designed to support patients with chronic conditions. These projects illustrate how computational health intelligence can translate biomedical insights into accessible, scalable digital health solutions.\n\t\t\t\t\t\t
Population and Environmental Health Intelligence<\/h4>\n\t\t\tHealth outcomes are influenced not only by biological and clinical factors but also by environmental and societal conditions. Computational Health Intelligence therefore extends to modelling population-level health dynamics using large-scale environmental and behavioural datasets.\n\u00a0\nCurrent work includes interdisciplinary research on\u00a0urban environmental factors and their influence on human health responses<\/b>, particularly within high-density cities. These studies aim to develop predictive frameworks that link environmental characteristics, human perception, and health outcomes, informing future urban design and public health strategies.\n
\n\t\tTranslational Research and Collaborative Impact\n\t<\/h4>\nA defining characteristic of Computational Health Intelligence is its strong emphasis on\u00a0translational impact and interdisciplinary collaboration<\/b>. Research in this area is conducted in close partnership with clinicians, healthcare institutions, and international collaborators to ensure that computational innovations are translated into real-world healthcare applications.\nThis work has been supported through numerous competitive research grants across domains including\u00a0digital health interventions, rehabilitation technologies, clinical informatics, and biomedical data science<\/b>. These projects involve collaborations with hospitals, healthcare organisations, and academic partners, enabling the development and evaluation of intelligent healthcare technologies in real clinical and community settings.\n
\n\t\tResearch Vision\n\t<\/h4>\nThe long-term vision of Computational Health Intelligence is to establish a\u00a0multiscale computational framework for healthcare<\/b>, where insights generated at the molecular and biological level inform clinical decision-making and personalised health interventions.\nBy integrating\u00a0bioinformatics, clinical analytics, and digital health innovation<\/b>, this research area aims to develop intelligent systems that not only predict disease but also support proactive and personalised approaches to maintaining health and well-being.\nThrough these efforts,\u00a0BioDigital Health Intelligence<\/b>\u00a0seeks to contribute to the development of\u00a0future healthcare systems that are data-driven, adaptive, and centred on improving human health outcomes.<\/b>\n
\n\t\tResearchers\n\t<\/h4>\n\t