A Genetic Algorithm-Based Ensemble Learning Framework for Drug Combination Prediction

doi:10.1021/acs.jcim.3c00260

A Genetic Algorithm-Based Ensemble Learning Framework for Drug Combination Prediction

2023-06-12

4

PRE

AI

DOI：10.1021/acs.jcim.3c00260

Original

Original request for help

Save

Abstract

En 中文

Combination therapy is a promising clinical treatmentstrategyfor cancer and other complex diseases. Multiple drugs can target multipleproteins and pathways, greatly improving the therapeutic effect andslowing down drug resistance. To narrow the search space of synergisticdrug combinations, many prediction models have been developed. However,drug combination datasets always have the characteristics of classimbalance. Synergistic drug combinations receive the most attentionin clinical application but are in small numbers. To predict synergisticdrug combinations in different cancer cell lines, in this study, wepropose a genetic algorithm-based ensemble learning framework, GA-DRUG,to address the problems of class imbalance and high dimensionalityof input data. The cell-line-specific gene expression profiles underdrug perturbations are used to train GA-DRUG, which contains imbalanceddata processing and the search of global optimal solutions. Comparedto 11 state-of-the-art algorithms, GA-DRUG achieves the best performanceand significantly improves the prediction performance in the minorityclass (Synergy). The ensemble framework can effectively correct theclassification results of a single classifier. In addition, the cellularproliferation experiment performed on several previously unexploreddrug combinations further confirms the predictive ability of GA-DRUG.

Keywords:

GENERATION

DESIGN

A Genetic Algorithm-Based Ensemble Learning Framework for Drug Combination Prediction

Abstract

Journal

Organization