Exploring key gene in Acute myeloid leukemia with weighted genes co-expression network analysis

Purpose: This study applied the WGCNA method to calculate the data of AML acquired from the TCGA database to discover the genes that are highly related to AML, and figure out key genes from these highly expressed genes, providing novel potential targets and biomarkers for therapy.Methods: The TCGA-AML was downloaded to construct gene co-expression module according to the correlation of genes, and the most significant module exhibiting the highest correlation coefficiency with clinical characteristics was selected. The Cox proportional hazards model was selected for gene survival analysis to screen risk factors related to AML death. Time-dependent ROC was analyzed to screen the genes with the largest ROC curve area and identified as the key genes, and Cox model forest and Nomogram method were constructed to verify the prognosis model.Results: According to the correlation of gene expression, 22 co-expression modules were found. The Grey60 module was significantly positively correlated with clinical characteristics (Cor=0.27, P<0.001). PTCRA, SLC4A3, SERPINF1, PPM1, and MB were screened by survival analysis. Furthermore, PPM1J with the largest AUC area was selected as the key gene. The death risk of the PPM1J high expression group was 2.83 times higher than that of the PPM1J low expression group. The survival rates of patients with different ages and cytogenetic risk types were statistically significantly different (P <0.05). Nomogram displayed that the expression of PPM1J in patients has a great impact on the survival time. The predicted 1-year, 3-year, and 5-year survival rates of patients in the low-expression group of PPM1J are >85%, >85%, and >70%, respectively, while the 1-year, 3-year, and 5-year survival rates of patients in the high-expression group are significantly decreased to 65%, 35%, and 20%, respectively.Conclusion: In this study, the PPM1J gene was screened out by constructing a gene co-expression network, which may become a potential genetic biomarker, affecting the occurrence and prognosis of AML.