Objective: In this study, clinical bioinformatics analysis was used to identify potential biomarkers of acute myeloid leukemia (AML) occurrence and development, drug resistance, and poor prognosis to provide a theoretical basis for the treatment of AML.
Methods: On the basis of the TCGA, GEO, and GTEx databases, an AML secondary database was established, and differential expression analysis and WGCNA were carried out to identify genes related to the prognosis of AML patients. Survival analysis was carried out for internal verification of key genes, and GEO data were used for external verification to obtain core genes related to prognosis. For differentially expressed genes, the EpiMed platform independently developed by the team was used for drug prediction.
Results: A total of 36 overlapping genes were obtained via difference analysis and WGCNA. Enrichment analysis revealed that the overlapping genes were associated with neutrophil activation, transcription dysregulation, AML, apoptosis, and other biological indicators. A protein interaction network was constructed for NCOA4, ACSL4, DPP4, ATL1, MT1G, ALOX15, and SLC7A11, which are key genes. Survival analysis revealed that NCOA4, ACSL4, DPP4, and ATL1 significantly affected the survival of patients with AML. The GSE142698 dataset verified that MPO, BCL2A1, and STMN1 had a statistically significant impact on the survival of AML patients.
Conclusion: NCOA4, ACSL4, DPP4, and ATL1 may be potential biomarkers related to the survival and prognosis of patients with AML, and the calcineurin signaling pathway is associated with the risk of vascular fragility in AML patients, which can provide a reference for further research and optimization of treatment regimens.
Keywords: acute myeloid leukemia; clinical biological information; coexpression; network analysis; nomogram prognosis.
© 2024 John Wiley & Sons Ltd.