Molecules to medicine: advancing integrated computational frameworks for biomedical applications
Loading...
Files
Full Text E-thesis
Zip File
Date
2025
Authors
Mi, Yanlin
Journal Title
Journal ISSN
Volume Title
Publisher
University College Cork
Published Version
Abstract
With the increasing complexity of biomedical knowledge and its interdisciplinary nature, effective translation of knowledge faces significant challenges. From molecular discovery to therapeutic application, it is challenging to ensure conceptual continuity and operational consistency due to the fragmented and domain-specific nature of current technology. In this thesis, we propose a unified computational framework, Translation-as-Infrastructure, to systematically support the stable migration of biomedical knowledge across cognitive levels, data forms, and operating systems.
Guided by the principles of semantic fidelity, structural unity, and output interpretability, the framework addresses four major translational bottlenecks: translating patient narratives into structured system logic, translating heterogeneous clinical data into trustworthy diagnostic reasoning, translating molecular insights into clinically actionable knowledge, and translating protein properties into scalable, computable representations. To operationalize this framework, some specific systems and computational tools have been developed and validated, including patient service platforms (MSaaVS, PHPlace), data modeling systems (Strengthening Diagnostic Translation), cross-level inference models (EPOP, AI-driven protein function analysis), and a modular protein computing infrastructure (PROFASA, PS-GO, Silver Surfer, ProteinFlow).
By integrating these systems, the thesis constructs a continuous, multi-layered translational pipeline from molecules to medicine. While this work represents an initial foundational effort, primarily focusing on structured knowledge migration rather than fully addressing semantic transparency, causal inference, and dynamic system self-adaptation, it provides a scalable architecture for future development of adaptive biomedical knowledge systems. By rethinking translation as a computational and infrastructural mechanism, this study develops a new paradigm for multidisciplinary knowledge engineering, providing methodological improvements and practical tools to assist the next generation of biomedical informatics and translational research.
Description
Keywords
Protein design , Pelvic organ prolapse , Machine learning , Extracellular matrix biomarkers , Surgical outcome prediction , Telemedicine , Software-as-a-service , Healthcare accessibility , Computational biology , Molecular visualisation , Data preprocessing
Citation
Mi, Y. 2025. Molecules to medicine: advancing integrated computational frameworks for biomedical applications. PhD Thesis, University College Cork.