Integrating multi-omics data with deep learning for predictive modelling of autoimmune and neurodegenerative diseases

Gopal, Priya; Vellaipandiyan, Solaimurugan; Navi, Nazi Mohammed

Abstract

In precision medicine, predicting the onset and progression of autoimmune and neurodegenerative diseases poses a multifaceted challenge due to their intricate molecular complexities. Traditional diagnostic methods, while foundational, often fall short in capturing the full scope of these dynamic and heterogeneous pathologies. This study introduces an advanced framework that integrates multi-omics data, including genomics, transcriptomics, proteomics, and metabolomics-to enhance predictive modeling for such conditions. Leveraging the power of sophisticated deep learning methodology allows the framework to establish patterns and interdependencies at a scale greater than trivial for several layers of biological strata. The Feedforward model within this framework demonstrates exceptional predictive performance, achieving an accuracy of 97.34%, an F1 score of 97.20%, and a balanced accuracy of 97.45%. The incorporation of explainable AI (XAI) techniques ensures model interpretability, facilitating the identification of novel biomarkers and revealing previously obscured pathways of disease progression. This integrative approach not only advances the precision of prognostic models but also catalyzes the development of personalized diagnostic and therapeutic strategies, marking a significant milestone in the evolution of precision medicine.

• Presentation Conference Type: Conference Paper (unpublished)
• Conference Name: AISB-2025
• Start Date: Jan 14, 2025
• End Date: Jan 16, 2025
• Acceptance Date: Jan 6, 2025
• Deposit Date: Feb 20, 2025
• Peer Reviewed: Peer Reviewed
• Keywords: Multi-Omics Integration; Deep Learning; Autoimmune Diseases; Neurodegenerative Disorders; Explainable AI; Predictive Modeling
• Public URL: https://uwe-repository.worktribe.com/output/13780049