Gastric cancer remains a highly prevalent malignancy worldwide with its molecular features poorly understood. To gain full insight into its genetic landscape, we performed whole-transcriptome sequencing on human tumors and adjacent non-tumors to predict the function of microRNA, long coding RNA, circular RNA, and mRNA, as well as estimate their correlation with gastric cancer characteristics through construction of ceRNA, WGCNA and PPI network. Functional enrichment analysis annotated nucleic acid binding, enzyme activity and binding related to differentially expressed miRNAs (dif-miRNAs); energy binding and enzyme binding related to dif-lncRNAs; protein binding and enzyme activity related to dif-circRNAs; protein digestion and absorption related to dif-mRNAs. The expression of key miR-135a-5p, lncRNAs-MSTRG.48856.1, ENST00000569981, MSTRG.22826.1, ENST00000564492, circRNAs-CCSER2, FNDC3B, CORO1C, FAM214A were validated by real-time PCR. The ceRNA network filtered 14 miRNAs, 30 lncRNAs, and 6 mRNA in the lncRNA-ceRNA axis and 8 miRNAs, 9 circRNAs, and 3 mRNA in the circRNA-ceRNA axis. Genes involved in ceRNA were annotated to be closely related to tumor material synthesis and metabolism. The WGCNA network filtered gene clusters related to TNM traits and extracted the hub genes CLDN10, CD177, newGene_35523, newGene_51201, CEACAM7, and newGene_46634. These genes were associated with cell proliferation, metabolism, and enzyme activity regulation. The PPI network analyzed the stable interaction relationships of the hub genes. Our research provides a valuable resource for understanding the molecular mechanisms of gastric cancer from the perspective of tumor metabolism.