Peripheral blood TCR clonotype diversity as a biomarker for immune-mediated diseases 2777

Abstract Description AI has become an indispensable tool in immunology, utilizing advanced computational and machine learning techniques to analyze complex immunological data, unravel immune system mechanisms, and accelerate the discovery of novel diagnostics and therapeutics. Our work is to explore the potential of AI in immune-related diseases diagnostics. Colorectal cancer (CRC) ranks among the top three most commonly diagnosed cancers globally and continues to be a significant contributor to cancer-related deaths. Early diagnosis of CRC is crucial for ensuring appropriate treatment, as the cancer can metastasize to other organs. However, current diagnostic approaches such as colonoscopy, fecal occult test and imaging technology have notable limitations, such as the risk of perforation, false-positive results, and missed non-bleeding lesions. Therefore, in this study, we proposed a novel approach for CRC diagnosis from the perspective of immune cells combined with machine learning methods, which is less invasive and more effective. This diagnostic model demonstrated robust performance in both the test set and external validation cohort, achieving AUC values of 0.98 and 0.94, respectively. Furthermore, we successfully identified the top 20 biomarkers that contributed most to CRC diagnosis. This model can also be extended for the diagnosis of other immune-mediated diseases. Funding Sources This work was supported by the Institute Digital Medicine grant at the City University of Hong Kong. Topic Categories Computational and Systems Immunology (COMP)