Identification and Validation of Metabolic Hub Genes Using WGCNA and Prognostic Risk Modeling in Acute Myeloid...

Area of Science:

• Hematology
• Molecular Biology
• Cancer Research

Background:

• Acute myeloid leukemia (AML) is an aggressive cancer with poor outcomes, especially in older adults.
• Metabolic reprogramming is increasingly recognized for its role in AML progression, immune evasion, and treatment resistance.
• The prognostic impact of metabolism-related molecular networks in AML requires further elucidation.

Purpose of the Study:

• To identify key metabolism-related genes driving AML progression.
• To develop a reliable prognostic signature for predicting patient survival in AML.
• To explore the link between metabolic alterations and immune responses in AML.

Main Methods:

• Integrated transcriptomic analysis of multiple AML datasets (GEO, TCGA-LAML).
• Weighted gene co-expression network analysis (WGCNA) to identify AML-associated gene modules.
• Protein-protein interaction, functional enrichment, and LASSO Cox regression for prognostic model development.

Main Results:

• A four-gene metabolic signature (CYP4F3, PFKL, G6PD, DNMT3A) emerged as an independent predictor of overall survival in AML.
• High-risk patients identified by the signature exhibited significantly poorer survival (p < 0.0001).
• The model demonstrated robust and consistent predictive performance across validation cohorts, highlighting altered glycolysis and redox homeostasis.

Conclusions:

• A novel metabolic-based prognostic signature offers strong predictive utility for AML risk stratification.
• Identified metabolic hub genes offer insights into metabolic dysregulation, immune remodeling, and potential therapeutic targets in AML.
• This signature can improve patient management by identifying high-risk individuals needing closer monitoring or intensified treatment.