Construction of a solid Cox model for AML patients based on multiomics bioinformatic analysis

Area of Science:

• Hematology
• Oncology
• Immunology

Background:

• Acute myeloid leukemia (AML) is a complex blood cancer.
• The bone marrow microenvironment significantly influences AML progression and treatment resistance.
• Identifying reliable prognostic biomarkers is crucial for AML patient management.

Purpose of the Study:

• To discover immune-related biomarkers for predicting acute myeloid leukemia (AML) prognosis.
• To develop a multiomics-based prognostic model for AML.
• To investigate the role of immune cell infiltration in AML patient survival.

Main Methods:

• Utilized The Cancer Genome Atlas (TCGA) and GSE106291 datasets for multiomics analysis.
• Constructed a LASSO-Cox regression model to identify prognostic genes.
• Performed molecular docking and single-cell RNA sequencing (scRNA-seq) for validation and immune cell analysis.

Main Results:

• Identified distinct AML subtypes based on immune infiltration (IL type vs. IR type), with IL type showing better prognosis.
• Developed a 5-hub gene classifier (ADAMTS3, CD52, CLCN5, HAL, ICAM3) for AML prognosis prediction.
• Validated the prognostic model using independent datasets and clinical samples.

Conclusions:

• Integrated multiomics analysis provides a robust approach for AML prognostic biomarker discovery.
• The developed 5-gene classifier can predict overall survival in AML patients.
• This prognostic model may serve as a valuable therapeutic classifier for guiding AML treatment strategies.