"Integrating Machine Learning and Complex Networks for Biological and Clinical Data Analysis"

The thesis defense of David G. Aragonés will take place on Tuesday, May 28, 2024. The dissertation is entitled: “Integrating Machine Learning and Complex Networks for Biological and Clinical Data Analysis", and has been supervised by Gabriel F. Calvo. His thesis defense committee is constituted by Maria Alieva (Institute of Biomedical Research Sols-Morreale, Madrid), Fátima Al-Shahrour (National Center for Cancer Research, Madrid), and Salvador Chulián (University of Cádiz).

This thesis presents a comprehensive investigation into metabolic biomarkers of diabetes, dimensionality reduction techniques, and the dynamic behavior of cells in inflammation response. First, it introduces labile hemoglobin as a biomarker for detecting short-term glucose variations, utilizing nonlinear dimensionality reduction to identify distinct subclusters within participant groups. Subsequently, it addresses variable selection in biological datasets, proposing a new method using correlation networks' bootstrapping tailored for nonlinear dimensionality reduction. Finally, this methodology is applied to high-dimensional biological datasets from live imaging of leukocytes in inflamed tissues, revealing insights into cell behavior and identifying the kinase Fgr as a key driver of pathogenic neutrophil states.