Michael P. Menden, Dennis Wang, Mike J. Mason, Bence Szalai, Krishna C. Bulusu, Yuanfang Guan, Thomas Yu, Jaewoo Kang, Minji Jeon, Russ Wolfinger, Tin Nguyen, Mikhail Zaslavskiy, AstraZeneca-Sanger Drug Combination DREAM Consortium, In Sock Jang, Zara Ghazoui, Mehmet Eren Ahsen, Robert Vogel, Elias Chaibub Neto, Thea Norman, Eric K. Y. Tang, Mathew J. Garnett, Giovanni Y. Di Veroli, Stephen Fawell, Gustavo Stolovitzky, Justin Guinney, Jonathan R. Dry, and Julio Saez-Rodriguez
The effectiveness of most cancer targeted therapies is short-lived. Tumors often develop resistance that might be overcome with drug combinations. However, the number of possible combinations is vast, necessitating data-driven approaches to find optimal patient-specific treatments. Here we report AstraZeneca’s large drug combination dataset, consisting of 11,576 experiments from 910 combinations across 85 molecularly characterized cancer cell lines, and results of a DREAM Challenge to evaluate computational strategies for predicting synergistic drug pairs and biomarkers. 160 teams participated to provide a comprehensive methodological development and benchmarking. Winning methods incorporate prior knowledge of drug-target interactions. Synergy is predicted with an accuracy matching biological replicates for >60% of combinations. However, 20% of drug combinations are poorly predicted by all methods. Genomic rationale for synergy predictions are identified, including ADAM17 inhibitor antagonism when combined with PIK3CB/D inhibition contrasting to synergy when combined with other PI3K-pathway inhibitors in PIK3CA mutant cells.
By evaluating predictions from a large number of teams, we were able to discern important strategies for predicting drug synergy from molecular and chemical traits. As with most DREAM Challenges, we observed that the machine learning method itself has little impact on overall performance. Aggressive pre-filtering that incorporates clean sparse network data to consider feature relevance to drug targets and cancer was successfully used by top performers to limit model complexity and improve model generalizability. Despite the complexity of the problem, many teams reached the upper-bound of performance levels based on variability in experimental replicates. This was further confirmed when top-performing models were applied to an independent dataset, demonstrating robustness to assay variability, and context heterogeneity.
Nature Communications, 10, 2674, 2019-06-17.