Simultaneous Optimization of Drug Combination Dose‐Ratio Sequence with Innovative Design and Active Learning

Abstract

AbstractStrategies for identifying potential drug combinations provide great benefits for the clinical treatment of complex diseases. However, existing approaches primarily focus on optimization of drug combinations and doses without comprehensive joint consideration of drug sequence, which compresses the therapeutic space. Herein, a novel combinatorial drug screening technique is proposed, named innovative design and active learning (IDEAL). This approach integrates innovative design, experimentation, and active learning. IDEAL allows for accurate predictions on the most efficacious drug combination, drug doses, and drug sequence with minimal amount of experimental effort. In this study, IDEAL is applied to lymphoma treatment, which successfully identifies the best drug‐dose sequence combination of chemotherapeutics on Raji cells. The IDEAL optimized regimen indicates that both drug dose and sequence are critical to ensure the optimal phenotype efficacy. This research offers an efficient methodology to identify the optimal drug administration setups in combination therapy.