Identification of drug combinations on the basis of machine learning to maximize anti-aging effects

Area of Science:

• Biogerontology
• Computational Biology
• Genetics

Background:

• Aging is a complex process influenced by numerous genetic factors, making the identification of effective anti-aging agents difficult.
• Understanding age-associated genetic changes is crucial for developing targeted anti-aging strategies.
• Current in silico drug repositioning methods have limitations in predicting effective drug combinations for aging.

Purpose of the Study:

• To develop an improved machine learning system for identifying potential anti-aging drug combinations.
• To investigate the efficacy of combinational drug treatments in reversing aging effects.
• To enhance the prediction accuracy of drug repositioning strategies for anti-aging applications.

Main Methods:

• Utilized a machine learning system, specifically a deep-neural network (DNN), trained on aging-related gene expression patterns.
• Modified existing expression profile methods with a stepwise-filtered approach to improve prediction.
• Applied the system to identify combinational drug pairs targeting differential expressed genes (DEG) in leukemia cells.

Main Results:

• Combinational drug pairs demonstrated a significantly enhanced anti-aging effect compared to single drug treatments.
• The modified machine learning approach improved the prediction success rate for anti-aging agents.
• Identified specific drug combinations with higher efficacy in retarding aging effects on leukemia cells.

Conclusions:

• Innovative machine learning techniques can effectively identify potent anti-aging drug combinations.
• Targeting differential expressed genes with drug pairs offers a promising strategy for combating aging.
• This study provides a foundation for developing more efficacious anti-aging therapies.