BackgroundNatural killer (NK) cells play a significant role in anti-tumor immunity, and their involvement has been documented in various cancers. However, a deeper understanding of the mechanisms by which NK cells influence gastric cancer progression remains necessary. MethodsWe utilized the Cancer Genome Atlas (TCGA) database to acquire transcriptional profiles, clinical information, and mutation data for gastric cancer patients. R software and associated packages were employed for all analyses of this publicly available data. ResultsWe used multiple algorithms to evaluate the tumor microenvironment in gastric cancer samples. We performed differential expression analysis to pinpoint genes related to NK cells. Utilizing this data, we developed a prognostic model featuring three crucial NK cell-related genes: MAB21L2, ARPP21, and MUCL1. This model showed strong predictive performance in the training and validation groups. Consistently, patients identified as high-risk according to our model had worse overall survival rates. To further elucidate the biological differences between high-risk and low-risk patients, we performed enrichment analyses focusing on biological pathways and immune-related factors. Additionally, we observed a correlation between higher risk scores and non-responsiveness to treatment. Interestingly, high-risk patients were found to be potentially more sensitive to axitinib. We selected MUCL1 for further investigation due to its potential role in the model. While MUCL1 mRNA levels were elevated in both gastric cancer and paired normal tissues, protein expression analysis using the Human Protein Atlas database revealed a decrease in MUCL1 protein levels within tumor tissues. ConclusionsOur findings contribute to a more comprehensive understanding of the role of NK cells in gastric cancer and highlight MUCL1 as a promising therapeutic target.