Integrated bioinformatic analysis and machine learning developed a prognostic model based on mitochondrial function...

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Acute myeloid leukemia (AML) presents a significant global health challenge.
• Mitochondria are crucial in cancer development, influencing energy, biosynthesis, and signaling.
• The prognostic and therapeutic roles of mitochondrial function in AML require further exploration.

Purpose of the Study:

• To identify key mitochondrial genes associated with AML prognosis.
• To develop a predictive model for AML risk stratification.
• To investigate the impact of mitochondrial genes on immune cells in AML.

Main Methods:

• Integrated mitochondrial gene expression with bulk RNA sequencing.
• Employed 14 machine learning algorithms to identify predictive gene combinations.
• Developed a MitoScore and a nomogram incorporating clinical variables.

Main Results:

• MitoScore effectively classified AML samples into low-risk and high-risk groups.
• High-risk group showed worse prognosis, enriched immune-related pathways, and higher mutation frequency.
• Observed specific mitochondrial gene expression patterns in T cell subsets and AML development stages.

Conclusions:

• The MitoScore framework offers a robust tool for AML prognostic assessment.
• MitoScore facilitates personalized treatment planning for acute myeloid leukemia patients.
• Highlights the potential of targeting mitochondrial function for AML therapy.