Related Experiment Video
Updated: Jul 4, 2025
09:53
Quantifying the Brain Metastatic Tumor Micro-Environment using an Organ-On-A Chip 3D Model, Machine Learning, and Confocal Tomography
Published on: August 16, 2020
7.2K
A machine learning model identifies M3-like subtype in AML based on PML/RARα targets.
Tingting Shao1, Jianing Li1, Minghai Su1
1College of Bioinformatics Science and Technology, Harbin Medical University, Harbin, Heilongjiang Province 150001, China.
Iscience
|February 7, 2024
Summary
A new machine-learning model identifies Acute Myeloid Leukemia (AML) M3 subtype and similar patients. This approach may expand eligibility for ATRA/ATO therapy, improving outcomes for more AML patients.
Area of Science:
- Hematology
- Genomics
- Computational Biology
Background:
- Acute Myeloid Leukemia (AML) M3 subtype is characterized by the PML/RARα fusion event.
- Current standard treatment for M3 AML involves ATRA/ATO combination therapy.
Purpose of the Study:
- Develop a machine-learning model to accurately identify M3 AML patients based on gene expression.
- Identify 'M3-like' non-M3 AML patients who could potentially benefit from ATRA/ATO therapy.
Main Methods:
- Analysis of gene expression data from 1228 AML patients.
- Machine learning model development focusing on PML/RARα target gene expression.
- Identification of M3-like patients using myeloid progenitor cell proportions and gene expression patterns.
Main Results:
- The developed machine-learning model achieved high accuracy in identifying M3 AML patients.
- M3-like patients were identified, exhibiting strong common myeloid progenitor (GMP) features and distinct genomic profiles.
- M3-like patients demonstrated lower immune activity and better clinical survival rates.
Conclusions:
- Machine learning can effectively identify M3 AML and M3-like patients.
- Identifying M3-like patients may broaden the application of ATRA/ATO therapy.
- This research deepens the understanding of AML pathogenesis and treatment strategies.
