Abstract
Gastric cancer remains a leading cause of cancer-related deaths with considerable heterogeneity among patients. Accurate classifications play a pivotal role in prognosis prediction and personalized therapeutic strategies. Considering the practicality of typing and its closer clinical relevance, in this study, we leveraged multi-omics data, specifically transcriptomics RNA-sequencing (mRNA) and DNA methylation data from the TCGA-STAD cohort, for clustering analysis. The integration of multi-omics data concerning prognosis facilitated cluster analysis through the implementation of ten clustering algorithms. A total of 359 gastric cancer (GC) samples were collected and categorized. Moreover, external validation datasets from diverse sequencing technologies corroborated the robustness of the clustering model. The relationships between the separate subgroups and clinical pathological characteristics, immune infiltration characteristics, immune checkpoint, genomic mutation and so on were meticulously examined. Meanwhile, potential responses to immunotherapy and chemotherapy were also assessed to enhance the clinical applicability of the molecular subtypes. Three subtypes (CS1, CS2, and CS3) were identified for gastric cancer, and exhibiting distinct prognostic status, activation of cancer-related pathways, TME compositions, immune checkpoints, sensitivity to chemotherapy and immunotherapy.