Search algorithms as a framework for the optimization of drug combinations

Area of Science:

• Biomedical research
• Computational biology
• Pharmacology

Background:

• Effective treatment of complex diseases often requires combination therapies.
• Current methods for identifying optimal drug combinations rely heavily on empirical clinical experience.
• There is a need for more systematic and efficient approaches to discover novel therapeutic combinations.

Purpose of the Study:

• To introduce and evaluate a novel application of modified search algorithms for optimizing therapeutic intervention combinations.
• To demonstrate the efficacy of these algorithms in identifying effective drug combinations across different biological models.
• To highlight the potential of information theory-based search strategies in accelerating drug discovery.

Main Methods:

• Adapted search algorithms, originally from digital communication, to optimize combinations of therapeutic interventions.
• Applied algorithms in biological experiments involving Drosophila melanogaster to restore age-related decline in heart function and exercise capacity.
• Utilized algorithms in experiments with human cancer cells to identify selective drug combinations.
• Validated algorithm performance through simulations using a network model of cell death.

Main Results:

• Search algorithms identified optimal drug combinations using significantly fewer tests compared to fully factorial searches.
• Algorithms achieved a highly significant enrichment of selective drug combinations for killing human cancer cells.
• Simulations showed algorithms identifying optimal combinations in 80-90% of tests, outperforming random searches (15-30%).

Conclusions:

• Modified search algorithms from information theory offer a powerful tool for enhancing the discovery of novel therapeutic drug combinations.
• This interdisciplinary approach has the potential to revolutionize how combination therapies are developed.
• The findings suggest a general strategy for solving complex biomedical problems through computational methods.