Blood biomarkers reveal progression of neurodegeneration using transcriptomic and proteomic datasets

• Main Author: Lim, Louis Li Jie
• Other Authors: -
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Computer and Information Science; Medicine, Health and Life Sciences; Alzheimer's disease; Transcriptomics, proteomics; Randomforest; XGBoost
• Online Access: https://hdl.handle.net/10356/180648
• Institution: Nanyang Technological University

The use of blood biomarkers in Alzheimer’s disease (AD) diagnosis has been touted as a non invasive alternative to current forms of diagnosis. However, whilst potential biomarkers associated to AD have been found, there has not been much research done for earlier stages of neurodegeneration, such as mild cognitive impairment (MCI). The heterogenicity of MCI and AD due to its multifaceted nature prevents any direct correlation of specific genes or proteins to effectively diagnose the stage of disease progression. Our study used a supervised machine learning model, xgboost, to reduce multi-modal data into linear regression for analysis, and enriched any differentially altered genes or proteins via publicly available gene annotation libraries, to identify the altered pathways in the progression of AD. We identified three key pathways altered in different stages of neurodegeneration: chronic inflammation, vascular damage and cholesterol dysregulation. These pathways show a potential temporal relationship, disambiguating the heterogenicity of MCI and AD as a progressive disease, which may be beneficial for future research and development of early diagnosis techniques for MCI and AD.