Gastric cancer (GC) is one of the leading causes of malignancy‑associated mortality worldwide. However, the underlying molecular mechanisms of GC are unclear and the prognosis of GC is poor. Therefore, it is important and urgent to explore the underlying mechanisms and screen for novel diagnostic and prognostic biomarkers, as well as therapeutic targets. In the current study, scale‑free gene co‑expression networks were constructed using weighted gene co‑expression network analysis, the potential associations between gene sets and clinical features were investigated, and the hub genes were identified. The gene expression profiles of GSE38749 were downloaded from the Gene Expression Omnibus database. RNA‑seq and clinical data for GC from The Cancer Genome Atlas were utilized for verification. Furthermore, the expression of candidate biomarkers in gastric tissues was investigated. Survival analysis was performed using Kaplan‑Meier and log‑rank test. The predictive role of candidate biomarkers in GC was evaluated using a receiver operator characteristic (ROC) curve. Gene Ontology, gene set enrichment analysis and gene set variation analysis methods were used to interpret the function of candidate biomarkers in GC. A total of 29 modules were identified via the average linkage hierarchical clustering. A significant module consisting of 48 genes associated with clinical traits was found; three genes with high connectivity in the clinical significant module were identified as hub genes. Among them, SLC5A6 and microfibril‑associated protein 2 (MFAP2) were negatively associated with the overall survival, and their expression was elevated in GC compared with non‑tumor tissues. Additionally, ROC curves indicated that SLC5A6 and MFAP2 showed a good diagnostic power in discriminating cancerous from normal tissues. SLC5A6 and MFAP2 were identified as novel diagnostic and prognostic biomarkers in GC patients; both of these genes were first reported here in connection with GC and deserved further research.