Brief Summary: Gastric cancer continues to have a poor prognosis primarily due to the inability to detect it in its early stages. This study will develop and validate a blood assay to facilitate the non-invasive detection of gastric cancer.

Detailed Description: Gastric cancer continues to have a poor prognosis primarily due to the inability to detect it in its early stages. Because conventional endoscopy is invasive and costly, gastric cancer is currently not considered to be screenable at a population level. However, if one could find less invasive and cheaper tools that accurately detect gastric cancer in its early stages, it could make a significant difference. Accurate biomarkers could help identify patients with gastric cancer before it becomes incurable. This study aims to develop a non-invasive test to detect gastric cancer early. It consists of four phases: 1. Discovering potential biomarkers with a comprehensive and genome-wide transcriptomic sequencing analysis that will involve gastric cancer tissue, normal tissue, and serum samples from patients with gastric cancer, as well as samples from people without the disease. 2. Using machine learning to develop a combination "signature" of cell-free (cf) and exosomal (exo)-miRNA in serum specimens from a training cohort. 3. A validation of this signature in an independent cohort to confirm its accuracy. 4. An evaluation of the temporal trend of this signature in paired samples collected pre-surgery and post-surgery to investigate their potential and specificity as indicators of minimal residual disease. In summary, this study aims to develop a highly accurate and cost-effective blood test for detecting gastric cancer early. Success could lead to significant improvements in clinical practice by catching cancer when it is most treatable. By combining different genetic markers (cell-free microRNA and exosomal microRNA) for accuracy, this study has the potential to reduce gastric cancer deaths and could lead to new screening methods in the future.