ABSTRACT Head and neck cancer is one of the most prevalent aggressive cancers worldwide, with the majority being squamous cell carcinoma (SCC), or HNSCC. The well-known risk factors associated with the genesis of HNSCC include human papillomavirus and environment factors including tobacco and alcohol use. Despite more than two decades of research into the molecular genetics of HNSCC, there is still lack of prognostic biomarkers that can guide current treatment and management of HNSCC patients. Previous studies and our preliminary analysis showed that top immune-centric biomarkers had the most promising prognostic value for HNSCC patients compared to other immunotherapy-favorable cancers such as skin and lung cancers. In this R01 project, we propose a large-scale secondary analysis of existing transcriptomic and genetic data to elucidate the prognostic role of intrinsic immune evasion mechanisms in head and neck cancer. The underlying central hypothesis is that tumor-intrinsic and host-intrinsic immune signaling co-modulate the effectiveness of existing therapies in HNSCC; and elucidation of the interaction between targeted genomic/genetic alternations and background tumor-immune signatures via efficient statistical and machine learning methods can lead to novel prognostic biomarkers, as well as potential therapeutic targets. To this end, we propose three specifics aims. In the first aim, we will generate HNSCC-specific immune signatures by integrating single-cell and bulk-tumor transcriptomic data using novel machine learning methods developed by our team. In the second aim, we will discover and prioritize immune-related long non-coding RNA (lncRNA) biomarkers by recalling existing transcriptome data and applying an efficient genome-wide screening method. In the third aim, we will combine baseline germline variants with tumor immune signatures and further examine its prognostic value in patients treated by immune checkpoint blockade (provide by multiple institutes including Moffitt, OSU and UNC). Collectively, we propose the most extensive study to date in HNSCC investigating under-investigated element of the genomics source for predicting clinical and tumor immune phenotypes. The new omics data resource and resulted biomarkers, especially the immune-related biomarkers, from our research have potential to impact upon the diagnosis and prognosis of HNSCC patients and could guide more efficient and personalized immunotherapy selection in the future.