AMLRS

AMLRS computes an AML risk score from a panel of 10 survival-related genes to stratify acute myeloid leukemia patients by overall survival using survival analysis methods.

Key Features:

• Risk Stratification: Employs log-rank tests, univariate COX regression, and LASSO-COX to categorize AML patients into high-risk and low-risk groups.
• AML Risk Score (AMLRS): Derives a risk score from a panel of 10 survival-related genes that acts as an independent prognostic factor across multiple datasets.
• Data Utilization: Developed using 1707 samples from three public databases split into meta-training, meta-testing, and validation sets.
• Validation and Performance: Validated across datasets with low-risk patients showing significantly longer overall survival (P < .001) and time-dependent ROC AUCs ranging from 0.5854 to 0.8066 for 1-, 3-, and 5-year survival predictions.
• Integration of Clinical Parameters: Incorporates two clinical parameters into a nomogram to improve predictive performance for 1-, 3-, and 5-year overall survival.

Scientific Applications:

• Prognostic Evaluation: Provides a risk stratification framework and an independent prognostic score for AML overall survival to support clinical risk assessment.
• Research Tool: Supports investigation of genetic determinants of AML survival and facilitates exploration of targeted therapies.

Methodology:

Data from three public databases (1707 samples) were divided into meta-training, meta-testing, and validation sets; log-rank tests, univariate COX regression, and LASSO-COX were used to identify survival-related genes and construct the AMLRS; model performance was evaluated using time-dependent ROC curves and validation across multiple datasets.