Predictive approaches for drug combination discovery in cancer

Area of Science:

• Computational biology
• Pharmacology
• Oncology

Background:

• Drug combinations offer a promising strategy to enhance cancer therapy efficacy and overcome drug resistance by targeting multiple disease pathways.
• Current drug combinations are often discovered empirically, leading to unclear molecular mechanisms and limiting rational therapeutic design.
• High-throughput screening for drug combinations is infeasible due to the vast number of potential pairings.

Purpose of the Study:

• To review and categorize state-of-the-art computational approaches for predicting efficacious drug combinations in cancer.
• To elaborate on the limitations and challenges associated with current computational drug combination prediction methods.
• To discuss the significance of recent pan-cancer drug combination datasets for method development and refinement.

Main Methods:

• Systematic literature review and categorization of computational methods for drug combination prediction.
• Analysis of existing data sources and their impact on predictive model performance.
• Discussion of challenges in computational drug combination discovery, including lack of consensus and methodological standardization.

Main Results:

• A wide array of computational methods, techniques, and hypotheses have been developed for drug combination prediction.
• The field currently lacks a clear consensus on preferred computational approaches and comprehensive data categorization.
• Recent advancements in pan-cancer drug combination datasets are crucial for advancing predictive accuracy and developing more robust methods.

Conclusions:

• There is a critical need for systematic and rational computational approaches to guide the discovery of effective cancer drug combinations.
• Further development and standardization of computational methods, coupled with the utilization of comprehensive datasets, are essential for improving drug combination prediction.
• Addressing the current limitations and challenges will accelerate the translation of computational predictions into clinical applications for enhanced cancer therapy.