2026-09-22T05:09:50Z https://www.ideals.illinois.edu/oai-pmh

oai:www.ideals.illinois.edu:2142/132547 2026-02-20 col_2142_5131 col_2142_14795 com_2142_5130 com_2142_14794 com_2142_14793 com_2142_8903

application/pdf en text Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo terms The student, Yiquan Wang, accepted the attached license on 2025-12-01 at 00:43. The student, Yiquan Wang, submitted this Dissertation for approval on 2025-12-01 at 01:02. This Dissertation was approved for publication on 2025-12-02 at 09:03. DSpace SAF Submission Ingestion Package generated from Vireo submission #22987 on 2026-02-19 at 18:25:45 Integrative approaches to decipher influenza evolution, antibody responses, and AI-driven specificity prediction Wang, Yiquan 2025-12-02 Wu, Nicholas Wu, Nicholas Brooke, Christopher Stadtmueller, Beth Tajkhorshid, Emad Deep learning Antibody Recognition Influenza Virus Artificial Intelligence eng The ongoing threat of viral pathogens, such as SARS-CoV-2 and influenza Viruses, highlights the urgent need to understand immune responses and viral evolution to guide therapeutic and vaccine development. This dissertation integrates high-throughput experimental techniques and artificial intelligence (AI) to address key questions in virus-immunity interactions through three interconnected research areas: (1) deep mutational scanning (DMS) to map sequence–function relationships in influenza viral proteins; (2) large-scale analysis of antibody responses to SARS-CoV-2 and influenza; and (3) development of AI models to predict antibody specificity. Chapter 1 introduces the rapid advancement of high-throughput and AI methodologies for studying immune responses and viral evolution. Chapter 2 presents a robust DMS pipeline that reveals high N-terminal tolerance in the nuclear export protein (NEP) and identifies charge-driven epistasis as a constraint on neuraminidase (NA) antigenic evolution. Chapter 3 describes large-scale profiling of antibody repertoires across viral pathogens, identifying critical residues in IGHV1-69 broadly neutralizing antibodies that target the hemagglutinin (HA) stem. Chapter 4 showcases AI-driven models for predicting antibody specificity and highlights their promise for therapeutic design. Chapter 5 synthesizes key findings and outlines future directions. By combining high-throughput experimentation with AI, this dissertation advances our understanding of host–pathogen interactions and provides new tools for vaccine design and immunotherapy. 2025-12 Thesis https://hdl.handle.net/2142/132547 © 2025 Yiquan Wang. All rights reserved Biochemistry Biochemistry University of Illinois Urbana-Champaign Ph.D. Dissertation