MACHINE LEARNING-DRIVEN INTEGRATIVE TRANSCRIPTOMIC ANALYSIS TO IDENTIFY SHARED BIOMARKERS AND MOLECULAR PATHWAYS BETWEEN HUNTINGTON’S DISEASE AND COMMON VARIABLE IMMUNODEFICIENCY

Abstract

It is becoming more widely acknowledged that neurodegenerative illnesses are intricate systemic conditions impacted by immunological dysregulation, extracellular matrix remodelling, and modified molecular signalling beyond the central nervous system. This study investigated the relationship between immunological dysfunction and neurodegeneration using a combined Machine Learning (ML) and literature-based paradigm. To find common biomarkers associated with disease, preprocessing, normalisation, supervised machine learning, and Explainable Artificial Intelligence (xAI) were used to examine public transcriptome datasets from Huntington's disease (HD) and Common Variable Immunodeficiency (CVID). Support Vector Machine (SVM) performed better than the other classifiers, with an accuracy of 95.7% on the test set. While functional enrichment analysis identified pathways such as antigen processing and presentation, clathrin-mediated endocytosis, hyaluronan uptake and degradation, and axon guidance, SHAP-based feature ranking identified genes like CSF1, ITGB5, TCP11L2, CDC5L, L3MBTL2, and BTG2 as important contributors to classification, indicating a strong confluence between immune regulation and neuronal biology. Additionally, network analysis revealed that CSF1 and BTG2 might serve as significant regulatory hubs in miRNA–mRNA interactions. Simultaneously, hyaluronan was investigated as a biologically significant extracellular matrix component that affects the stability of the neuronal milieu, immune cell trafficking, and inflammation. According to the reviewed data, hyaluronan metabolism is extremely context-dependent as fragmented hyaluronan can stimulate inflammatory signalling and immunological activation, high-molecular-weight hyaluronan often maintains tissue homeostasis. Due to its impact on glial activation, perineuronal net remodelling, demyelination, and delayed regeneration, dysregulated hyaluronan turnover has been linked to neurodegenerative disease. All of these results point to a single paradigm that explains how immune dysregulation and neurodegeneration interact through shared transcriptional signatures and hyaluronan-associated pathways. Potential biomarkers and molecular pathways that could be helpful for upcoming diagnostic and immunologically guided treatment approaches were found in this investigation.