Bioinformatics Of Poor Prognostic Markers In Acute Myeloid Leukemia Analysis And Preliminary Verification

Objective: To analyze the transcriptomic data of patients with acute myeloid leukemia(AML)in the gene chip database by clinical bioinformatics to obtain genes that are highly expressed in AML and associated with poor prognosis,and to verify their expression differences in AML by fluorescence quantitative PCR.Methods:(1)The GSE9476 and GSE37642 datasets were selected from the gene expression omnibus(GEO)database,then we screened the expression-related differential genes in the dataset GSE9476 and the genes associated with poor prognosis of AML in GSE37642,and the above results were intersected to obtain the target genes.(2)Univariate and multifactorial independent prognostic analyses were performed using R software,and correlation analyses of age and molecular characteristics were performed to investigate the relationship between target gene expression levels and age,gene mutations,and fusion genes.(3)The samples were divided into two groups of high and low expression according to the median value of the target gene expression level,and the genes that were differentially expressed between the two groups were analyzed for functional enrichment analysis,and the top 20 genes were selected to analyze the strength of their correlation between two groups.(4)The genes differentially expressed between the two groups were analyzed using the String website,and the protein interaction network was mapped.(5)The Oncomine and gene expression profilling interactive analysis(GEPIA)databases were used to verify the differences in expression levels of target genes in AML patients and normal samples.(6)Fluorescent quantitative PCR was applied to detect the difference in expression of target genes in AML patients and normal subjects.Results: Screening yielded two target genes,SHLD2 and H1-0,both highly expressed in AML patients and associated with poor prognosis in AML,among which SHLD2 could be an independent poor prognostic factor in AML,and the expression level of SHLD2 was lower in patients with RUNX1-RUNX1T1;H1-0 could not be an independent prognostic factor,but patients older than or equal to 60 years of age or with RUNX1 mutations had higher levels of H1-0 expression.Fluorescence quantitative PCR showed that SHLD2 expression levels were lower in AML samples than in normal human samples,and H1-0 expression levels were higher in AML samples than in normal human samples,but were not statistically significant.Conclusions: Using bioinformatics to study microarray data of AML patients,we screened and validated two highly expressed genes associated with poor prognosis,SHLD2 and H1-0,respectively,in which SHLD2 could be an independent poor prognostic factor for AML and could be a potential marker and therapeutic target for the development of AML.