Prediction of NR3C1 as a methylation marker for the prevention and treatment of gastric cancer in Rhizoma Atractylodis Macrocephalae based on machine learning algorithm and bioinformatics analysis.

Objective: The aim of this study was to identify the markers of Rhizoma Atractylodis Macrocephalae (RAM) for the prevention and treatment of gastric cancer using bioinformatics analysis.

Methods: The main active components of RAM were screened using the Traditional Chinese Medicine Systematic Pharmacology Profiling Platform (TCMSP) and SwissADME, the target genes of RAM were screened using WGCNA and three machine learning algorithms, and the target genes were analyzed clinically and by methylation.

Results: Three core genes, namely, CA2, HSP90AA1, and NR3C1, were screened by WGCNA and three machine learning algorithms. Clinical correlation analysis and epigenetic analyses showed that these genes play the most important role in gastric cancer. In gastric cancer, there was a strong correlation between NR3C1 methylation and its mRNA expression, suggesting that methylation of NR3C1 may be involved in the regulation of its expression. Therefore, methylation correlation analysis of NR3C1 was performed, and it was found that the methylation type of NR3C1 was mainly m6A methylation; the frequency of methylation hypermutation in the CDS region was also high, and the homologous region and the promoter of the NR3C1 gene showed hypomethylation and hypermethylation differences in the analyses of gastric tissues, races, and gastric cancer subtypes, respectively. Among the methylated peninsular mutations, the TSS200; 5'UTR, 5'UTR, and TSS1500; 5'UTR regions were statistically significant.

Conclusion: NR3C1 can be used as a potential methylation marker of RAM for the prevention and treatment of gastric cancer.